Coronary artery model reconstruction method based on coronary artery angiography

Abstract

The invention relates to the technical field of model reconstruction, in particular to a method for reconstructing a coronary artery model based on coronary artery angiography. The method comprises the following steps: acquiring a coronary artery scanning image of a patient; performing three-dimensional coronary artery space structure construction on the coronary artery image of the patient to generate a three-dimensional coronary artery space model; performing blood vessel diameter analysis on the three-dimensional coronary artery space model to generate coronary artery blood vessel diameter data; acquiring an X-ray angiography image of a patient; performing coronary artery vein structure construction on the X-ray angiography image of the patient to generate a three-dimensional coronary artery vein model; and performing blood vessel wall thickness analysis on the three-dimensional coronary artery vein model to generate coronary artery blood vessel wall thickness data. According to the method, multi-modal modeling is carried out by combining the coronary artery scanning image and the X-ray angiography image of the patient, anomaly analysis is carried out on the coronary artery, and the accuracy and reliability of coronary artery model reconstruction are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of model reconstruction, and in particular to a method for reconstructing a coronary artery model based on coronary angiography. Background Art

[0002] With the advancement of computer vision and image processing algorithms, 3D reconstruction technology based on coronary angiography has undergone a revolutionary transformation. The introduction of machine learning and deep learning technologies has enabled the automatic extraction of 3D coronary artery structure from 2D images, significantly reducing the need for manual work. For example, algorithms such as convolutional neural networks (CNNs) can automatically identify vascular boundaries from vascular images and generate accurate 3D models. In recent years, deep learning-based generative adversarial networks (GANs) have also emerged in the field of medical image processing. These techniques are capable of generating high-quality 3D coronary artery models from limited data. GAN models can learn from a large number of coronary artery samples and then generate realistic new samples, addressing data shortages and improving reconstruction accuracy and stability. However, traditional modeling often only reconstructs vascular structure without considering important parameters such as vessel diameter and wall thickness. Furthermore, they rely solely on single-modality data, such as coronary angiography, resulting in low accuracy and reliability in model reconstruction. Summary of the Invention

[0003] Based on this, it is necessary to provide a method for reconstructing a coronary artery model based on coronary angiography to solve at least one of the above technical problems.

[0004] To achieve the above object, a method for reconstructing a coronary artery model based on coronary angiography is provided, the method comprising the following steps:

[0005] Step S1: Acquire a coronary artery scan image of the patient; construct a three-dimensional coronary artery spatial structure on the patient's coronary artery image to generate a three-dimensional coronary artery spatial model; perform vascular diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vascular diameter data;

[0006] Step S2: Acquire an X-ray angiography image of the patient; construct a coronary artery vascular structure based on the X-ray angiography image of the patient to generate a three-dimensional coronary artery vascular model; perform a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery vascular wall thickness data;

[0007] Step S3: Constructing a coronary artery flow structure based on the coronary artery wall thickness data and the coronary artery diameter data to generate a three-dimensional coronary artery blood flow model; integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing coronary artery abnormality analysis on the coronary artery integrated model to generate deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas;

[0008] Step S4: reconstruct the coronary artery model of the coronary artery integrated model using the deep exploration data of the highly abnormal part and the shallow exploration data of the low abnormal part to generate a coronary artery lesion model; predict the effect of virtual surgery on the coronary artery lesion model to construct a corresponding implementation plan template.

[0009] The present invention generates an accurate 3D coronary artery spatial model by constructing a 3D spatial structure from a patient's coronary artery scan image. This helps doctors gain a more comprehensive understanding of the patient's arterial anatomy. By analyzing vessel diameter and wall thickness, detailed coronary artery vascular characteristics are provided, facilitating assessment of vascular health and potential pathological conditions. A 3D coronary blood flow model is constructed based on the vessel wall thickness and diameter data. This helps doctors understand blood flow within the arteries and assess the impact of hemodynamics on vascular health. The 3D coronary artery spatial model, 3D coronary artery venation model, and 3D coronary blood flow model are integrated into an integrated coronary artery model. This integration provides more comprehensive and integrated information on coronary artery structure and function, providing a foundation for subsequent analysis and diagnosis. The generated deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas help doctors locate and analyze coronary artery abnormalities. This data is then used to reconstruct lesions within the integrated coronary artery model, generating a coronary artery lesion model. This helps doctors more accurately diagnose coronary artery lesions. Predicting the outcome of virtual surgery using the coronary artery lesion model helps doctors develop appropriate implementation plan templates. This prediction can be simulated before surgery, helping doctors evaluate the effectiveness of different treatment options and improving the success rate and safety of the surgery. Therefore, the present invention combines the patient's coronary artery scan images and X-ray angiography images for multimodal modeling, and analyzes coronary artery abnormalities, thereby improving the accuracy and reliability of coronary artery model reconstruction.

[0010] The present invention provides the following advantages: by processing a patient's coronary artery scan images, a three-dimensional coronary artery spatial model can be constructed and vessel diameter analysis can be performed. This helps doctors understand the morphology, structure, and vessel diameter of the patient's coronary arteries. By processing the patient's X-ray angiography images, a three-dimensional coronary artery vascular model can be constructed and vessel wall thickness analysis can be performed. This helps doctors understand the vascular structure and wall thickness of the patient's coronary arteries. Based on the coronary artery wall thickness data and coronary artery diameter data, a three-dimensional coronary artery blood flow model can be constructed. This facilitates simulation and analysis of blood flow within the patient's coronary arteries. By integrating the three-dimensional coronary artery spatial model, the coronary artery vascular model, and the coronary artery blood flow model, an integrated coronary artery model is generated. By performing anomaly analysis on the integrated coronary artery model, deep exploration data for highly abnormal areas and shallow exploration data for less abnormal areas can be determined, thereby providing more accurate information on coronary artery lesions. Using the deep exploration data for highly abnormal areas and the shallow exploration data for less abnormal areas, the integrated coronary artery model is reconstructed to generate a coronary artery lesion model. The coronary artery lesion model can then be used to predict the effectiveness of virtual surgery, assessing the effectiveness of different treatment options and possible corrective surgery plans for coronary artery lesions. Therefore, the present invention improves the accuracy and reliability of coronary artery model reconstruction by combining a patient's coronary artery scan images and X-ray angiography images for multimodal modeling and performing coronary artery abnormality analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A schematic flow chart of the steps of a method for reconstructing a coronary artery model based on coronary angiography;

[0012] Figure 2 for Figure 1 Detailed implementation steps of step S3 in FIG.

[0013] Figure 3 for Figure 2 Detailed implementation steps of step S31 are shown in the flowchart;

[0014] Figure 4 for Figure 2 Detailed implementation steps of step S33 are shown in the flowchart.

[0015] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0016] The following is a clear and complete description of the technical method of the present invention in conjunction with the accompanying drawings. Obviously, the embodiments described are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present invention.

[0017] In addition, the accompanying drawings are merely schematic illustrations of the present invention and are not necessarily drawn to scale. Identical reference numerals in the figures denote identical or similar parts, and thus repetitive descriptions thereof will be omitted. Some of the block diagrams shown in the accompanying drawings are functional entities that do not necessarily correspond to physically or logically separate entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor and / or microcontroller approaches.

[0018] It should be understood that although the terms "first," "second," and the like may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used solely to distinguish one element from another. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element, without departing from the scope of the exemplary embodiments. The term "and / or" as used herein includes any and all combinations of one or more of the listed associated items.

[0019] To achieve this, please refer to Figures 1 to 4 A method for reconstructing a coronary artery model based on coronary angiography, the method comprising the following steps:

[0020] Step S1: Acquire a coronary artery scan image of the patient; construct a three-dimensional coronary artery spatial structure on the patient's coronary artery image to generate a three-dimensional coronary artery spatial model; perform vascular diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vascular diameter data;

[0021] Step S2: Acquire an X-ray angiography image of the patient; construct a coronary artery vascular structure based on the X-ray angiography image of the patient to generate a three-dimensional coronary artery vascular model; perform a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery vascular wall thickness data;

[0022] Step S3: Constructing a coronary artery flow structure based on the coronary artery wall thickness data and the coronary artery diameter data to generate a three-dimensional coronary artery blood flow model; integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing coronary artery abnormality analysis on the coronary artery integrated model to generate deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas;

[0023] Step S4: reconstruct the coronary artery model of the coronary artery integrated model using the deep exploration data of the highly abnormal part and the shallow exploration data of the low abnormal part to generate a coronary artery lesion model; predict the effect of virtual surgery on the coronary artery lesion model to construct a corresponding implementation plan template.

[0024] The present invention generates an accurate 3D coronary artery spatial model by constructing a 3D spatial structure from a patient's coronary artery scan image. This helps doctors gain a more comprehensive understanding of the patient's arterial anatomy. By analyzing vessel diameter and wall thickness, detailed coronary artery vascular characteristics are provided, facilitating assessment of vascular health and potential pathological conditions. A 3D coronary blood flow model is constructed based on the vessel wall thickness and diameter data. This helps doctors understand blood flow within the arteries and assess the impact of hemodynamics on vascular health. The 3D coronary artery spatial model, 3D coronary artery venation model, and 3D coronary blood flow model are integrated into an integrated coronary artery model. This integration provides more comprehensive and integrated information on coronary artery structure and function, providing a foundation for subsequent analysis and diagnosis. The generated deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas help doctors locate and analyze coronary artery abnormalities. This data is then used to reconstruct lesions within the integrated coronary artery model, generating a coronary artery lesion model. This helps doctors more accurately diagnose coronary artery lesions. Predicting the outcome of virtual surgery using the coronary artery lesion model helps doctors develop appropriate implementation plan templates. This prediction can be simulated before surgery, helping doctors evaluate the effectiveness of different treatment options and improving the success rate and safety of the surgery. Therefore, the present invention combines the patient's coronary artery scan images and X-ray angiography images for multimodal modeling, and analyzes coronary artery abnormalities, thereby improving the accuracy and reliability of coronary artery model reconstruction.

[0025] In the embodiment of the present invention, reference Figure 1 The above is a schematic flow chart of the steps of a method for reconstructing a coronary artery model based on coronary angiography according to the present invention. In this example, the method for reconstructing a coronary artery model based on coronary angiography includes the following steps:

[0026] Step S1: Acquire a coronary artery scan image of the patient; construct a three-dimensional coronary artery spatial structure on the patient's coronary artery image to generate a three-dimensional coronary artery spatial model; perform vascular diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vascular diameter data;

[0027] In an embodiment of the present invention, a coronary artery scan is performed on a patient using a medical imaging device, such as computed tomography (CT) or magnetic resonance imaging (MRI), to obtain coronary artery image data. Image processing and segmentation techniques are then applied to the coronary artery images to extract information about the coronary artery vascular structure. This may include using image processing algorithms to perform operations such as edge detection, threshold segmentation, and region growing to obtain the three-dimensional spatial structure of the coronary arteries. The extracted coronary artery vascular structure information is converted into a three-dimensional model representation. Computer graphics and three-dimensional reconstruction techniques can be used to convert the two-dimensional image data into a three-dimensional coronary artery spatial model. This may include methods such as surface reconstruction and voxelization to construct a three-dimensional model with spatial position and morphological features. Vascular diameter analysis is performed on the three-dimensional coronary artery spatial model to obtain coronary artery diameter data. This can be achieved by extracting and measuring vascular branches from the coronary artery spatial model. Common methods include curvature- and width-based vascular branch tracking algorithms and morphological operations.

[0028] Step S2: Acquire an X-ray angiography image of the patient; construct a coronary artery vascular structure based on the X-ray angiography image of the patient to generate a three-dimensional coronary artery vascular model; perform a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery vascular wall thickness data;

[0029] In an embodiment of the present invention, coronary angiography is performed on a patient using X-ray angiography to obtain angiographic image data of the coronary arteries. Image processing and segmentation techniques are then used to construct the coronary artery vascular network structure based on the obtained angiographic images. This may involve preprocessing steps such as noise removal and image enhancement. A vascular segmentation algorithm, such as threshold segmentation or edge detection, is then applied to extract the coronary artery vascular network structure. The extracted coronary artery vascular network structure is converted into a three-dimensional model for representation. A volume rendering algorithm, such as voxelization or surface reconstruction, can be used to convert the two-dimensional vascular network structure into a model with three-dimensional spatial information. The generated three-dimensional coronary artery network model is subjected to vascular wall thickness analysis to obtain coronary artery wall thickness data. This can be achieved by measuring the distance between the endothelium and adventitia surfaces of the blood vessels. Common methods include vascular wall segmentation algorithms based on curvature or grayscale gradient changes, morphological operations, and the like.

[0030] Step S3: Constructing a coronary artery flow structure based on the coronary artery wall thickness data and the coronary artery diameter data to generate a three-dimensional coronary artery blood flow model; integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing coronary artery abnormality analysis on the coronary artery integrated model to generate deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas;

[0031] In an embodiment of the present invention, a coronary artery flow model can be constructed based on coronary artery wall thickness data and coronary artery diameter data. This involves combining a vascular venous model with a physical blood flow model. By modeling the vascular blood flow as a fluid dynamics model, the flow behavior of blood in the coronary artery can be simulated. The specific fluid dynamics model construction process includes using the Navier-Stokes equations as basic equations. This equation group includes the continuity equation and the momentum equation, which are used to describe the movement and flow of the fluid. Considering the non-Newtonian nature of blood, the commonly used models are the compressible Navier-Stokes equations and the non-Newtonian model of blood, such as the Carreau-Yasuda model. The geometric shape of the blood vessel, including the diameter, length, curvature, etc. of the blood vessel, is determined. The inlet boundary conditions, such as the flow rate or velocity at the inlet, are set. The outlet boundary conditions, specifically the pressure value or resistance, are set. For the boundary conditions of the blood vessel wall, the no-slip wall condition is usually used, that is, the blood flow moves along the blood vessel wall, taking into account the density and viscosity of the blood. The density of blood is usually 1060 kg / m 3Viscosity is usually described using models such as the Carreau-Yasuda model, and numerical solutions are performed using computational fluid dynamics (CFD) software, such as ANSYS Fluent and COMSOL Multiphysics. The model geometry, boundary conditions, and fluid properties are set in the software, and appropriate meshing and solver parameters are selected to perform simulations to obtain results such as blood flow velocity and pressure distribution. The simulation results are then compared with experimental data to verify the accuracy of the model. The coronary blood flow model is integrated with the previously generated coronary artery spatial model and coronary artery venous model to generate a coronary artery integrated model. This can be achieved by spatially aligning and fusing the data of the three models. The specific spatial alignment process involves extracting some feature points or feature structures from each model, such as key vascular bifurcation points, vascular wall features, etc., and matching these feature points to establish a correspondence between different models. Based on the feature matching results, a spatial transformation is performed to align the coordinate systems of the different models. This involves transformations such as rotation, translation, and scaling. For example, point-by-point matching algorithms, least-squares registration, and nonlinear registration are used to perform spatial transformations. The aligned coronary blood flow model, the coronary spatial model, and the coronary vascular network model are combined. This involves mapping blood flow data onto the geometry of the spatial model and mapping velocity, pressure, and other information from the blood flow model onto the vessel interior of the coronary spatial model. This is achieved by interpolating the flow model results onto the spatial model mesh to ensure consistency in geometry and flow characteristics between the combined models. The combined model is then validated, for example, by comparison with experimental data or clinical observations. The combined model includes coronary artery geometry, vascular network structure, and blood flow information. Coronary anomaly analysis is performed based on this combined coronary artery model. Computer-aided diagnosis algorithms, machine learning models, and other methods can be used to detect and quantitatively analyze coronary artery anomalies. This allows for the identification of high- and low-abnormality areas and the generation of corresponding deep and shallow exploration data. These data can provide quantitative indicators of coronary artery abnormalities, such as abnormal blood flow velocity and degree of stenosis.

[0032] Step S4: reconstruct the coronary artery model of the coronary artery integrated model using the deep exploration data of the highly abnormal part and the shallow exploration data of the low abnormal part to generate a coronary artery lesion model; predict the effect of virtual surgery on the coronary artery lesion model to construct a corresponding implementation plan template.

[0033] In an embodiment of the present invention, the coronary artery integrated model is reconstructed by utilizing the deep exploration data of the highly abnormal site and the shallow exploration data of the less abnormal site generated in step S3. This can be achieved by registering and fusing the morphological information of the abnormal site with the integrated model. The registration and fusion process can utilize image registration algorithms, three-dimensional reconstruction technology, and other methods to add the specific lesion details of the abnormal site to the integrated model to generate a coronary artery lesion model. Based on the coronary artery lesion model, virtual surgical effect prediction is performed. This can be achieved by using tools such as computer-assisted surgical planning software and simulation platforms to simulate and predict the effects of different surgical operations on the efficacy of coronary artery lesions. By adjusting the surgical plan and simulating the surgical operation, post-operative blood flow, vascular patency, and possible side effects can be predicted. This helps to formulate corresponding implementation plan templates and provide guidance and reference for real operations.

[0034] Preferably, step S1 includes the following steps:

[0035] Step S11: obtaining a coronary artery scan image of the patient;

[0036] Step S12: performing image denoising on the patient's coronary artery image to generate a coronary artery denoised image; performing image contrast enhancement on the coronary artery denoised image to generate a coronary artery enhanced image;

[0037] Step S13: performing image vascular region segmentation on the coronary artery enhanced image to generate a coronary artery vascular region segmentation image; performing image binarization on the coronary artery vascular region segmentation image to generate a coronary artery vascular region binarization image;

[0038] Step S14: constructing a three-dimensional coronary artery spatial structure based on the binarized image of the coronary artery region to generate a three-dimensional coronary artery spatial model; performing a vessel diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vessel diameter data.

[0039] The present invention can reduce noise and interference in coronary artery images and improve image clarity and contrast by denoising and contrast enhancement. This helps doctors observe and analyze the morphology and structure of coronary arteries more accurately. By performing vascular region segmentation on the coronary artery enhanced image, the vascular region in the image can be distinguished from other tissue regions. This helps to extract the shape and contour information of the coronary artery, providing an accurate basis for subsequent analysis and measurement. Based on the binary image of the coronary artery vascular region, a three-dimensional coronary artery spatial structure can be constructed. By constructing a three-dimensional model, a more comprehensive understanding of the three-dimensional structure, morphology, and path of the coronary artery can be obtained. This provides richer information for subsequent vascular analysis and diagnosis. By performing vascular diameter analysis on the three-dimensional coronary artery spatial model, the diameter information of the coronary artery can be measured. Vascular diameter is one of the key indicators for assessing the severity of coronary artery disease. It can provide important information such as vascular stenosis and the degree of lesions, helping doctors diagnose the disease and plan treatment.

[0040] In an embodiment of the present invention, the coronary artery image is denoised using an appropriate image denoising algorithm, such as a filter-based method, wavelet transform, or non-local averaging. The denoised image is contrast enhanced using histogram equalization, adaptive histogram equalization, or other enhancement algorithms. A suitable segmentation algorithm, such as threshold segmentation, watershed algorithm, or region growing, is used to segment the coronary artery vascular region from the image. A binarization operation is performed on the segmented vascular image to divide the vascular region into foreground (blood vessels) and background (other tissues), generating a binarized image of the coronary artery vascular region. Based on the binarized image of the coronary artery vascular region, a three-dimensional coronary artery spatial model can be constructed using techniques such as surface reconstruction and voxelization. To analyze the vascular diameter of the three-dimensional coronary artery spatial model, the coronary artery diameter data can be obtained by measuring parameters such as the cross-sectional diameter, length, and curvature of the blood vessels. The specific measurement method uses the medical imaging software OsiriX, opens the medical imaging file of the coronary artery, usually a file in DICOM format, and selects an appropriate measurement tool in the software, such as the line measurement tool or the circle measurement tool. On the cross-section of the coronary artery, use the tool to mark the two endpoints of the blood vessel or the edge of the blood vessel. For the diameter, the circle measurement tool is usually used. Place the two points of the tool on the inner wall or edge of the blood vessel. The software will automatically calculate the diameter. To measure the length, the line measurement tool can be used to measure the length of the blood vessel. By recording the measured diameter, length and other data, you can save it as a report or export it as a data file.

[0041] Preferably, step S14 includes the following steps:

[0042] Step S141: extracting the centerline of the coronary artery region from the binary image to obtain the centerline of the coronary artery region; skeletonizing the coronary artery region from the binary image based on the centerline to generate a coronary artery skeleton image;

[0043] Step S142: performing skeleton image pixel value change analysis on the coronary artery skeleton image to generate pixel change values in the direction of the vessel centerline; performing radial projection on the coronary artery skeleton image based on the pixel change values in the direction of the vessel centerline to generate radial projection data of the coronary artery main vessel;

[0044] Step S143: confirming the coronary artery position based on the coronary artery skeleton image to obtain coronary artery position data; performing vascular tortuosity analysis based on the coronary artery position data and the coronary artery main vessel radial projection data to generate coronary artery curvature data;

[0045] Step S144: Using the coronary artery curvature data, the coronary artery skeleton image is subjected to vascular stent detection to generate coronary artery branch point location data; based on the coronary artery branch point location data and the coronary main vessel radial projection data, a three-dimensional coronary artery spatial structure is constructed to generate a three-dimensional coronary artery spatial model; and the three-dimensional coronary artery spatial model is subjected to vascular diameter analysis to generate coronary artery diameter data.

[0046] The present invention extracts and skeletonizes the coronary artery region's binary image through vessel centerline extraction, generating a coronary artery centerline and vessel skeleton image. This facilitates further analysis of the vessel's morphology and structure. Pixel value analysis and radial projection of the vessel skeleton image yield pixel change values along the vessel centerline and radial projection data of the main vessel. This data helps measure vessel changes and morphological characteristics and provides information on the radial distribution of the vessels. Based on the vessel skeleton image, the location of the coronary arteries can be confirmed and vessel curvature analysis performed. Vascular curvature data reveals the degree of tortuosity and morphological characteristics of the vessels, playing an important role in assessing vascular health and pathology. Using this data to detect stents in the vessel skeleton image, branch points can be located. Based on branch point location data and radial projection data of the main vessel, a three-dimensional coronary artery spatial model can be constructed to further analyze the structure and connectivity of the coronary arteries. Vessel diameter analysis of the three-dimensional coronary artery spatial model yields coronary artery diameter data. This data is crucial for assessing stenosis severity, calculating hemodynamic parameters, and formulating treatment strategies.

[0047] In an embodiment of the present invention, by extracting the centerline of the coronary artery vascular region from a binary image of the coronary artery vascular region, image processing techniques such as edge detection and morphological operations can be used to extract the centerline of the coronary artery vascular region. Subsequently, the binary image is skeletonized according to the centerline of the coronary artery vascular region to obtain a coronary artery vascular skeleton image. The pixel value change analysis of the coronary artery vascular skeleton image can be performed mainly by calculating the change value of the pixel point along the direction of the vascular centerline to obtain the pixel change value in the direction of the vascular centerline, wherein the change value refers to the change in the vascular diameter along the direction of the vascular centerline. The specific steps are to extract the centerline of the coronary artery from the medical imaging data, and for each point on the centerline, calculate the vascular radius in a direction perpendicular to the centerline. By comparing the vascular radii of adjacent points, the diameter change of the vascular along the centerline can be calculated. The calculation process can be expressed mathematically as follows: Among them, ri+1 is the radius of the next point on the centerline, and ri-1 is the radius of the previous point on the centerline. In this way, the diameter change value of each point in the direction of the blood vessel centerline can be obtained. Subsequently, these change values can be used to perform radial projection on the coronary artery skeleton image to generate radial projection data of the coronary main vessel. Based on the coronary artery skeleton image, the blood vessel position is confirmed, and morphological operations and geometric shape analysis can be used to confirm the position of the blood vessel and obtain the coronary artery position data. According to the coronary artery position data and the coronary main vessel radial projection data, the blood vessel curvature analysis can be performed to calculate the curvature data of the coronary artery. The specific blood vessel curvature analysis involves measuring and calculating the curvature of the blood vessel path, which can be achieved through the coronary artery position data and the main vessel radial projection data. For the main vessel radial projection data, its curvature can be calculated using mathematical methods. A commonly used method is to use the three-point method to calculate the curvature through the positions of three adjacent points. From the curvature calculation, the curvature change of the blood vessel along the main vessel direction can be obtained. The curvature can usually be defined as the curvature change per unit length. It can be calculated using the following formula: Where Δκ represents the change in curvature, and ΔS represents the unit length. Curvature can indicate the degree of curvature near a specific point in a vessel. Using coronary artery curvature data, coronary artery skeleton images can be used for stent detection. Image segmentation algorithms and morphological operations can be used to detect stents and obtain coronary branch point location data. Based on this coronary branch point location data and radial projection data of the main coronary artery, a three-dimensional coronary artery spatial structure can be constructed, generating a three-dimensional coronary artery spatial model. For vessel diameter analysis of the three-dimensional coronary artery spatial model, measurement tools and algorithms can be used to calculate coronary artery diameter data. Specifically, in medical imaging software, vessel diameters can be measured directly on the image using a line or circle measurement tool. This involves selecting an appropriate measurement tool, such as a line or circle measurement tool, selecting a known length on the image as a dimension, and then measuring the vessel diameter using the measurement tool. Alternatively, automated measurement algorithms can be used to preprocess the image, such as edge detection, and then identify and measure the vessel diameters. The measured diameter data can be recorded and analyzed as needed, such as calculating the mean diameter or maximum diameter.

[0048] Preferably, step S2 includes the following steps:

[0049] Step S21: Acquire an X-ray angiography image of the patient;

[0050] Step S22: performing image preprocessing on the patient X-ray angiography image to generate a standard patient X-ray angiography image, wherein the image preprocessing includes image denoising, image brightness enhancement, and image edge sharpening;

[0051] Step S23: constructing the coronary artery vascular structure of the standard patient X-ray angiography image according to the coronary artery diameter data to generate a three-dimensional coronary artery vascular model;

[0052] Step S24: performing a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery wall thickness data.

[0053] The present invention obtains X-ray angiography images of patients, which is the basis for obtaining information about the patient's coronary artery vascular structure. X-ray angiography images can clearly display the position and morphology of blood vessels. The patient's X-ray angiography images are preprocessed, including denoising, brightness enhancement, and edge sharpening. These preprocessing operations can improve the quality and clarity of the image, making subsequent analysis and processing steps more accurate and reliable. The coronary artery vascular vascular structure is constructed based on the coronary artery diameter data to generate a three-dimensional coronary artery vascular model. Through this step, the two-dimensional vascular image can be converted into a three-dimensional vascular vascular model, which can provide a more detailed and comprehensive understanding of the structure and morphology of the coronary artery. The vascular wall thickness of the three-dimensional coronary artery vascular model is analyzed to generate coronary artery wall thickness data. This step can provide detailed information on the coronary artery wall, which is helpful for evaluating the health of the blood vessels and diagnosing and monitoring related diseases.

[0054] In an embodiment of the present invention, a patient is examined using X-ray angiography to obtain information about the vascular structure of the coronary arteries. This technology can clearly display the position and morphology of blood vessels by injecting a contrast agent and using X-rays for imaging. The acquired images are preprocessed to improve image quality and clarity to facilitate subsequent analysis and processing. Image preprocessing includes the following operations: using a denoising algorithm to remove noise from the image to improve image clarity. Adjusting the brightness and contrast of the image to make the features of the blood vessels more obvious. Enhancing the edge features of the image to make the blood vessel contours clearer. Based on the coronary artery diameter data, computer image processing technology is used to convert the standardized patient X-ray angiography image into a three-dimensional coronary artery vascular model. This process involves vessel segmentation, vessel connectivity, and vascular structure establishment to reconstruct the entire coronary artery network. Specifically, vessel segmentation is performed on the patient's X-ray angiography image. Based on the grayscale differences between the vessels and surrounding tissue, a threshold is set to separate the vessels from the background. The 2D image is then converted into 3D voxelized data, where each voxel represents a small cubic region within the image. Based on the voxelized data, a 3D vascular model is reconstructed using the voxel connectivity. Common algorithms include the Marching Cubes algorithm, which is used to convert the data into a 3D coronary artery vascular network model. The resulting 3D coronary artery vascular network model is then analyzed for vessel wall thickness. By measuring vessel wall thickness, coronary artery health can be assessed and relevant data can be obtained. This data can be used for the diagnosis and monitoring of coronary artery disease. The specific measurement process involves using an appropriate segmentation method to separate the vessel from the background, using an edge detection algorithm (such as the Canny edge detector) to locate the vessel wall edge, and calculating the distance from each point on the vessel wall to the vessel centerline. This distance is the wall thickness of the blood vessel. For the entire blood vessel segment, the wall thickness of each point is averaged to obtain the average wall thickness of the blood vessel. The blood vessel can be divided into several segments and the average wall thickness of each segment can be calculated separately.

[0055] Preferably, step S23 includes the following steps:

[0056] Step S231: hierarchically dividing the standard patient X-ray angiography image into angiographic images to obtain a surface angiography image and a deep angiography image; dividing the surface angiography image and the deep angiography image into overlapping regions to generate an angiography overlapping region image and an angiography non-overlapping region image;

[0057] Step S232: performing initial vascular choroid connection on the non-overlapping region images of the angiography according to the coronary artery diameter data to obtain an initial vascular choroid image;

[0058] Step S233: using the initial blood vessel venous image to predict the connection path of the overlapping region of the angiography overlapped region image, and generating blood vessel overlapped region connection path prediction data;

[0059] Step S234: Based on the vascular overlapping area connection path prediction data and the coronary artery vascular curvature data, the vascular connection of the angiography overlapping area image is verified to generate vascular connection verification result data; the coronary artery vascular vascular structure is constructed for the angiography overlapping area image and the angiography non-overlapping area image using the vascular connection verification result data to generate a three-dimensional coronary artery vascular model.

[0060] The present invention hierarchically divides standard patient X-ray angiography images to produce surface angiography images and deep angiography images. These images are then divided into overlapping regions to generate overlapping angiography images and non-overlapping angiography images. This helps decompose the angiography images into smaller regions, facilitating subsequent vascular connection analysis and path prediction. Using coronary artery diameter data, initial vascular choroidal connections are performed on the non-overlapping angiography images. This step aims to establish an initial vascular choroidal image, providing a foundation for subsequent connection path prediction and verification. Using this initial vascular choroidal image, connection path prediction is performed on the overlapping angiography images. By analyzing the morphology and structure of the vessels, the connection paths of the vessels in the overlapping region are predicted. This allows the direction and connection pattern of the vessels in the overlapping region to be inferred, providing a basis for subsequent verification. Based on the predicted connection path data for the overlapping region and the coronary artery curvature data, vascular connection verification is performed on the overlapping angiography images. By verifying the predicted connection paths, the accuracy and rationality of the predicted connection paths are assessed, thereby generating vascular connection verification result data. Based on these validation data, the coronary artery vascular structure was constructed from the angiography overlapping area images and the angiography non-overlapping area images to generate a 3D coronary artery vascular model. This model provides a detailed description of the coronary artery vascular structure, facilitating further analysis.

[0061] In an embodiment of the present invention, X-ray angiography images of standard patients are preprocessed, including denoising and enhancement. The preprocessed images are then hierarchically divided into surface angiography images and deep angiography images. The specific hierarchical division method uses a region growing method, starting with a seed point (a characteristic point of a surface or deep blood vessel) and using the region growing method to group adjacent pixels into the same category for image hierarchical division. The surface and deep angiography images are then divided into overlapping regions. Angiography overlapping region images and angiography non-overlapping region images are generated based on pixel values, edge detection, and other methods. Initial vascular network connections are performed on the non-overlapping region images based on coronary artery diameter data. Specifically, the coronary artery diameter data is converted into corresponding spatial coordinates. For non-overlapping vascular regions, the vessels are connected based on the diameter data to form vascular networks. Vascular connectivity is determined based on diameter size, with connections between vessels whose diameters are greater than a certain threshold being used for connection. Using image processing and analysis techniques, such as binarization and morphological operations, the initial vascular choroids are connected to form an initial vascular choroid image. The initial vascular choroid image is then used to predict the connection paths of the overlapping regions of the angiographic images. Using image segmentation and path planning algorithms, the overlapping regions of the angiographic images are analyzed to predict vascular connection paths and generate prediction data for the overlapping regions. Specifically, the overlapping regions are segmented to distinguish different vessels. For connected vascular regions, the different vessels are segmented using a region growing algorithm. If the boundaries between vessels are clear, the Canny edge detection method is used to find the vessel boundaries. Path planning is performed on the segmented regions to predict vascular connection paths. Shortest path algorithms, such as the Dijkstra algorithm and the A* algorithm, are used to search for the shortest path within the segmented regions. Based on the predicted connection path data and coronary artery curvature data, vascular connectivity is verified on the overlapping regions of the angiographic images. The accuracy of the predicted vascular connection paths is verified using methods such as vascular morphology analysis and curvature calculation, and vascular connection verification data is generated. Based on the vascular connection verification data, the coronary artery vascular network structure is constructed from the angiographic overlapping and non-overlapping region images. Using 3D modeling techniques, the segmented 2D image data is converted into 3D voxel data. Based on this voxel data, a 3D reconstruction algorithm, such as the Marching Cubes algorithm, is used to convert the voxel data into a 3D vascular model. The generated 3D vascular model is then optimized and smoothed to achieve a more realistic and continuous appearance. The verified vascular connection paths and vascular network structure are reconstructed to generate a 3D coronary artery vascular model.

[0062] Preferably, step S233 includes the following steps:

[0063] Step S2331: using the initial vascular choroid image to segment the vascular connection ports of the overlapping region image of the angiography to obtain the vascular connection port data of the overlapping region; dividing the vascular connection port data of the overlapping region into a data set to generate a model training set and a model test set;

[0064] Step S2332: performing model training on the model training set based on the support vector machine algorithm to generate a training model for the connection paths of the vascular overlapping regions; performing model testing on the training model for the connection paths of the vascular overlapping regions using the model test set to generate a prediction model for the connection paths of the vascular overlapping regions;

[0065] Step S2333: importing the overlapping region blood vessel connection port data into the blood vessel overlapping region connection path prediction model to perform overlapping region blood vessel connection path prediction, thereby generating overlapping region blood vessel connection path prediction data.

[0066] The present invention accurately extracts vascular connection port data within overlapping regions by segmenting vascular connection ports within overlapping angiographic images. This provides important input for subsequent path prediction. The overlapping region vascular connection port data is then partitioned into a model training set and a model test set. By training the training set using a support vector machine algorithm, a prediction model for the connection paths within overlapping regions can be constructed. By importing the overlapping region vascular connection port data into the trained overlapping region connection path prediction model, the connection paths within the overlapping regions can be predicted. This prediction process helps determine the connectivity relationships between vascular vascular networks and generates corresponding prediction data for the connection paths within the overlapping regions.

[0067] In an embodiment of the present invention, an image processing algorithm is used to preprocess the overlapping region images of angiography to remove interference and noise. A vascular connection port segmentation algorithm is applied to the preprocessed images to identify and extract the vascular connection ports in the overlapping region. The extracted vascular connection port data is divided into a data set, into a model training set and a model test set. This can be used for model training and evaluation. Using the data in the model training set as input, a support vector machine algorithm is used for model training. The algorithm learns the relationship between the vascular connection port data and the corresponding connection paths. The performance and accuracy of the trained model are evaluated by predicting the model test set data. This allows for the selection of optimal model parameters and optimization of model performance. The vascular connection port data in the overlapping region is then imported into the trained support vector machine model. The model will predict the vascular connection path of the overlapping area based on the input vascular connection port data, and generate vascular overlapping area connection path prediction data. The specific model construction includes preparing training set data and test set data, where the training set data includes the feature data of the vascular overlapping area and the corresponding connection path label, and the test set data also includes the feature data of the vascular overlapping area and the corresponding connection path label. The training set and the test set are divided into 70% for the model training set and 30% for the model test set. By selecting the kernel function of the SVM model, such as the linear kernel function, polynomial kernel function, Gaussian kernel function, etc., the choice of kernel function should be considered according to the characteristics of the data and the degree of nonlinearity. In many machine learning libraries, you can directly call the ready-made SVM model. Specifically, in Python, you can use sklear The SVC (support vector classifier) class in the n.svm library is used to build the model. The training set is used to train the model to allow the model to learn the relationship between the vascular connection port data and the connection path. Training is performed by calling the model's fit method and passing in the training data. After training is completed, the model performance needs to be evaluated and necessary adjustments need to be made. The prediction accuracy of the model is evaluated using the test set. If the amount of data allows, cross-validation can be performed to better evaluate the generalization ability of the model. The model parameters, such as regularization parameters and kernel function parameters, are adjusted based on the evaluation results to obtain better performance. The trained support vector machine model can be used to predict the connection path corresponding to new vascular connection port data. For new vascular connection port data, the predict method of the trained model is used for prediction, and the corresponding connection path data is generated based on the prediction results.

[0068] Preferably, step S3 includes the following steps:

[0069] Step S31: performing hemodynamic flow simulation based on the coronary artery wall thickness data and the coronary artery diameter data to generate coronary artery hemodynamic flow simulation data; constructing a coronary artery flow structure based on the coronary artery hemodynamic flow simulation data to generate a three-dimensional coronary artery blood flow model;

[0070] Step S32: integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing dynamic coronary angiography simulation on the coronary artery integrated model to generate a dynamic coronary angiography simulation video;

[0071] Step S33: Analyzing the arterial blood flow condition of the dynamic coronary angiography simulation video to generate arterial blood flow characteristic data; performing coronary artery abnormality analysis on the coronary artery integrated model based on the arterial blood flow characteristic data to obtain coronary artery abnormality analysis data;

[0072] Step S34: Compare the coronary artery abnormality analysis data with the preset standard abnormality threshold. When the coronary artery abnormality analysis data is greater than or equal to the preset standard abnormality threshold, a deep exploration of the coronary artery site of the patient is performed based on the coronary artery abnormality analysis data to obtain deep exploration data of the high abnormality site; when the coronary artery abnormality analysis data is less than the preset standard abnormality threshold, shallow exploration data of the low abnormality site is generated.

[0073] The present invention generates coronary hemodynamic flow simulation data by simulating blood flow in the coronary arteries and combining it with coronary artery arm thickness and diameter data. This helps to gain a deeper understanding of the flow characteristics and patterns of blood in the coronary arteries and provides foundational data for subsequent steps. An integrated coronary artery model can be generated by integrating a three-dimensional coronary artery spatial model, a coronary artery vascular model, and a blood flow model. The integration process requires preparing data for the three-dimensional coronary artery spatial model, the coronary artery vascular model, and the blood flow model. The coordinate systems of the three models are aligned to ensure they are in the same space. The data from the three models are then fused, typically combining the position, shape, and size of the vessels with dynamic blood flow information. The fusion method involves data superposition, integrating the aligned and fused data into a complete integrated coronary artery model. Based on the integrated data, a mathematical or computational model is established to describe the overall structure and hemodynamic characteristics of the coronary arteries. This integrated model can be used for simulations and analyses, such as simulating blood flow in the coronary arteries and analyzing information such as blood flow velocity and blood pressure distribution. Furthermore, dynamic coronary angiography simulations using this model can generate simulated videos, allowing doctors and researchers to intuitively observe and analyze coronary artery flow. Analysis of the simulated videos provides characteristic data about blood flow. Furthermore, using this characteristic data, the integrated coronary artery model is analyzed for coronary artery anomalies, generating coronary artery anomaly analysis data. This helps identify potential coronary artery anomalies and provides a basis for subsequent treatment and diagnosis. The coronary artery anomaly analysis data is compared with a set standard anomaly threshold. If the value is greater than or equal to the preset standard anomaly threshold, a deeper exploration of the coronary artery sites can be performed to obtain deep exploration data for highly abnormal sites. Conversely, if the value is lower than the preset standard anomaly threshold, shallow exploration data for less abnormal sites is generated. This data helps locate coronary artery anomalies and provides visual results, facilitating clinical diagnosis and coronary artery management decision-making.

[0074] As an example of the present invention, refer to Figure 2 As shown, in this example, step S3 includes:

[0075] Step S31: performing hemodynamic flow simulation based on the coronary artery wall thickness data and the coronary artery diameter data to generate coronary artery hemodynamic flow simulation data; constructing a coronary artery flow structure based on the coronary artery hemodynamic flow simulation data to generate a three-dimensional coronary artery blood flow model;

[0076] In an embodiment of the present invention, data on coronary artery wall thickness and diameter are collected. This data can be obtained using non-invasive imaging techniques such as computed tomography (CT) or magnetic resonance imaging (MRI). Computational fluid dynamics (CFD) methods are then used to simulate blood flow in the coronary arteries. CFD simulations use numerical methods and physical equations to simulate the flow behavior of blood in the vessels. Based on the coronary artery wall thickness and diameter data, a geometric model can be established, and the blood flow behavior can be simulated as mathematical equations. Using computer algorithms and numerical solution methods, simulated data on blood flow in the coronary arteries can be calculated. The data obtained from the hemodynamic simulations is used to construct a coronary artery flow structure. Based on the simulated data, the velocity distribution, pressure distribution, and other blood flow parameters in the coronary arteries can be determined. These parameters can be visualized using visualization techniques, visualizing the coronary artery blood flow structure as a three-dimensional model. Based on the results of the coronary artery flow structure construction, a three-dimensional coronary artery blood flow model can be generated. This model can be used in subsequent steps, such as model integration and coronary artery anomaly analysis.

[0077] Step S32: integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing dynamic coronary angiography simulation on the coronary artery integrated model to generate a dynamic coronary angiography simulation video;

[0078] In an embodiment of the present invention, a three-dimensional coronary artery spatial model, a three-dimensional coronary artery vascular model, and a three-dimensional coronary artery blood flow model are integrated. This can be accomplished using computer-aided design (CAD) software or three-dimensional modeling software. The data of the three models are merged and aligned to ensure that they are in the same spatial coordinate system. Dynamic coronary angiography simulation is performed using the integrated coronary artery model. Coronary angiography is an interventional examination used to assess the degree of stenosis and blood flow in the coronary arteries. In the simulation, the introduction of a contrast agent and its flow in the coronary arteries can be simulated and observed. This can be simulated using computational fluid dynamics (CFD) methods, taking into account the geometry of the blood vessels, blood flow parameters, and the transport behavior of the contrast agent. Based on the results of the coronary angiography simulation, a dynamic coronary angiography simulation video can be generated. Professional visualization software and technology are required to present the dynamic changes during the simulation process in video form. The video can display the blood flow and the transmission path of the contrast agent in the coronary arteries, simulating the actual coronary angiography process.

[0079] Step S33: Analyzing the arterial blood flow condition of the dynamic coronary angiography simulation video to generate arterial blood flow characteristic data; performing coronary artery abnormality analysis on the coronary artery integrated model based on the arterial blood flow characteristic data to obtain coronary artery abnormality analysis data;

[0080] In an embodiment of the present invention, dynamic coronary angiography simulation videos are analyzed to obtain blood flow characteristic data. This may involve calculating parameters such as blood flow velocity, blood pressure gradient, and vessel wall shear stress. This data can be obtained by post-processing the simulation videos using computational fluid dynamics (CFD) methods. CFD analysis is typically based on fluid dynamics equations and the boundary conditions of a coronary artery model. Blood flow characteristic data is extracted and calculated based on the coronary angiography simulation videos. This requires specialized analysis software and algorithms to process the simulation results, thereby obtaining parameters such as blood flow velocity, vascular resistance, and blood flow distribution. Arterial blood flow characteristic data is then used to analyze coronary artery anomalies within an integrated coronary artery model. By combining blood flow data with the coronary artery model, possible coronary artery anomalies, such as stenosis, obstruction, or thrombosis, can be detected and analyzed. This can be achieved by quantitatively analyzing the model and comparing the flow characteristic data with normal reference ranges. Based on the coronary anomaly analysis results, coronary anomaly analysis data is generated. This data may include information such as the degree of coronary artery stenosis, associated hemodynamic effects, and blood flow distribution in abnormal areas. The analysis results can be visualized using charts, reports, or other forms as needed.

[0081] Step S34: Compare the coronary artery abnormality analysis data with the preset standard abnormality threshold. When the coronary artery abnormality analysis data is greater than or equal to the preset standard abnormality threshold, a deep exploration of the coronary artery site of the patient is performed based on the coronary artery abnormality analysis data to obtain deep exploration data of the high abnormality site; when the coronary artery abnormality analysis data is less than the preset standard abnormality threshold, shallow exploration data of the low abnormality site is generated.

[0082] In an embodiment of the present invention, the coronary artery abnormality analysis data is compared with preset standard abnormality thresholds. These standard abnormality thresholds can be established based on clinical experience, research data, or guidelines and are used to determine coronary artery abnormalities. If the coronary artery abnormality analysis data is greater than or equal to the preset standard abnormality threshold, it indicates that the patient has a high coronary artery abnormality. In this case, a deep exploration of the patient's coronary artery site is required. This may include the use of coronary angiography, computed tomography (CT), nuclear medicine imaging, or other related technologies to further assess the abnormality of the coronary arteries. The results may include detailed information such as the degree of coronary artery stenosis, the location and extent of branch vessel lesions, etc. If the coronary artery abnormality analysis data is less than the preset standard abnormality threshold, it indicates that the patient has a low degree of coronary artery abnormality. In this case, deep coronary artery exploration is generally not required. Superficial exploration data of the low-abnormality site can be generated, including basic coronary artery structure, blood flow distribution, and coronary artery status within the normal range.

[0083] Preferably, step S31 includes the following steps:

[0084] Step S311: performing global external vascular feature screening on the three-dimensional coronary artery choroid model and the three-dimensional coronary artery spatial model according to the coronary artery wall thickness data and the coronary artery diameter data to obtain global external vascular feature data;

[0085] Step S312: performing a virtual endoscopy based on the global external vessel length data, the global external vessel diameter data, and the global external vessel tortuosity data, thereby obtaining global internal vessel characteristic data; setting blood flow boundary conditions based on the global external vessel characteristic data and the global internal vessel characteristic data, thereby obtaining vascular blood flow boundary condition data;

[0086] Step S313: Calculating the vascular flow velocity field using the vascular blood flow boundary condition data to generate blood flow velocity field data; analyzing the vascular pressure distribution based on the blood flow velocity field data and the vascular blood flow boundary condition data to generate coronary artery hemodynamic flow simulation data, wherein the coronary artery hemodynamic flow simulation data includes blood flow simulation data and blood flow shear stress simulation data;

[0087] Step S314: constructing the coronary artery blood flow structure using the blood flow simulation data and the blood flow shear stress simulation data to generate a three-dimensional coronary artery blood flow model.

[0088] The present invention obtains comprehensive external structural information about the coronary artery system by filtering and extracting global external vascular feature data based on coronary artery wall thickness data and coronary artery diameter data. This feature data can be used for subsequent internal structural analysis and simulation. Virtual endoscopy is performed based on the global external vascular length data, global external vascular diameter data, and global external vascular tortuosity data to obtain global internal vascular feature data. This step provides more detailed information about the internal structure of the coronary arteries, including lumen size, vessel tortuosity, and more. The global external and internal vascular feature data can be used to set vascular blood flow boundary condition data. These boundary conditions are key parameters for blood flow simulation, enabling simulation and analysis of blood flow within the coronary arteries. Blood flow velocity field calculations are performed using the vascular blood flow boundary condition data to generate blood flow velocity field data. Pressure distribution analysis is then performed using the blood flow velocity field data and vascular blood flow boundary condition data to generate coronary hemodynamic flow simulation data. This data provides quantitative and qualitative information about blood flow within the coronary arteries. Using the simulated blood flow volume and blood shear stress data, a three-dimensional coronary blood flow model can be constructed. This model can be used to further analyze coronary abnormalities, evaluate vascular lesions, predict hemodynamic effects, etc.

[0089] As an example of the present invention, refer to Figure 3 As shown, in this example, step S31 includes:

[0090] Step S311: performing global external vascular feature screening on the three-dimensional coronary artery choroid model and the three-dimensional coronary artery spatial model according to the coronary artery wall thickness data and the coronary artery diameter data to obtain global external vascular feature data;

[0091] In an embodiment of the present invention, data on coronary artery wall thickness and diameter are collected. This data can be obtained through measurement techniques, such as medical imaging (e.g., CT or MRI). Using the collected vascular data, a three-dimensional coronary artery network model and spatial model are constructed. This can be accomplished using computer-aided design and modeling software. An appropriate feature selection algorithm is selected, using the vascular wall thickness and diameter data as input to select vessels within the three-dimensional coronary artery network model and spatial model. The specific criteria for using this feature selection algorithm should consider relevance: features most relevant to vascular structure and characteristics are selected. For example, vascular wall thickness and diameter are related to vascular type, size, and shape; interpretability: features with biological significance are selected, taking into account the interpretability and explainability of the final application; and predictability: features that can effectively predict the target (here, vascular structure and shape, etc.) are selected. Common feature selection algorithms include threshold segmentation, morphological operations, and domain analysis. Based on the feature selection algorithm, vascular feature data, such as vessel length, curvature, and branching, are extracted and recorded. Simultaneously, the extracted feature data is processed and normalized for subsequent use. The extracted and processed global external vascular feature data is saved and output. This data can be used for subsequent internal structure analysis, blood flow simulation, etc.

[0092] Step S312: performing a virtual endoscopy based on the global external vessel length data, the global external vessel diameter data, and the global external vessel tortuosity data, thereby obtaining global internal vessel characteristic data; setting blood flow boundary conditions based on the global external vessel characteristic data and the global internal vessel characteristic data, thereby obtaining vascular blood flow boundary condition data;

[0093] In an embodiment of the present invention, a virtual endoscopy process is simulated based on global external vessel length data, global external vessel diameter data, and global external vessel tortuosity data. This can be achieved through computer simulation and image processing techniques. The specific steps include: establishing a virtual endoscopy model based on the global external vessel data and aligning it with the global external vessel data. Based on the characteristics of the virtual endoscope and the geometry of the global external vessel, an endoscopy path is planned to ensure smooth passage of the endoscope through the vessel. Endoscopic images are simulated along the endoscopy path. Ray tracing or other image synthesis techniques can be used to generate realistic endoscopic images. Based on the endoscopic images generated by virtual endoscopy, feature data of the global internal vessels is extracted. This feature data may include vessel diameter variations, vessel wall thickness, and vessel texture. Using the global external and internal vessel feature data, blood flow boundary conditions are set. This involves setting parameters such as blood flow velocity and vessel wall friction coefficient during the simulated blood flow process to simulate real-world blood flow conditions. The determined blood flow boundary condition data is saved and output. This data can be used for subsequent blood flow simulation and analysis.

[0094] Step S313: Calculating the vascular flow velocity field using the vascular blood flow boundary condition data to generate blood flow velocity field data; analyzing the vascular pressure distribution based on the blood flow velocity field data and the vascular blood flow boundary condition data to generate coronary artery hemodynamic flow simulation data, wherein the coronary artery hemodynamic flow simulation data includes blood flow simulation data and blood flow shear stress simulation data;

[0095] In an embodiment of the present invention, by utilizing vascular blood flow boundary condition data, numerical simulation methods (such as computational fluid dynamics simulation) can be used to calculate the vascular flow velocity field. Specific steps include: establishing a mathematical model of blood flow motion, taking into account the vascular geometry, fluid flow equations, fluid boundary conditions, etc. The vascular geometry is discretized into a grid so that calculations can be performed at each grid point. Numerical methods (such as the finite element method, the finite volume method, etc.) are used to discretize and solve the fluid flow equations to obtain vascular flow velocity field data. The blood flow velocity field data and vascular blood flow boundary condition data are used to analyze the vascular pressure distribution. This can be achieved by substituting the blood flow velocity field data and vascular blood flow boundary condition data into the fluid dynamics equations and solving them to obtain the pressure distribution within the blood vessel. Specific steps include: establishing the motion equations of blood flow in the blood vessel based on fluid dynamics theory. Applying the vascular blood flow boundary condition data to the fluid dynamics equations and setting the inlet and outlet boundary conditions. Discretizing and solving the fluid dynamics equations using numerical methods to obtain the pressure distribution data within the blood vessel. Based on blood velocity field data and vascular pressure distribution data, coronary artery hemodynamic flow simulation data can be generated, including blood flow rate simulation data and blood shear stress simulation data. By integrating the velocity distribution across the vascular cross section, blood flow distribution data at different locations can be obtained. Based on the blood velocity gradient and blood viscosity, the shear stress distribution data of blood flow within the vessel can be calculated.

[0096] Step S314: constructing the coronary artery blood flow structure using the blood flow simulation data and the blood flow shear stress simulation data to generate a three-dimensional coronary artery blood flow model.

[0097] In an embodiment of the present invention, vascular geometry information of the coronary arteries is acquired using vascular imaging techniques (such as CT scans and MRI). Based on the blood flow distribution data and blood shear stress data, the vascular geometry is reconstructed using computer-aided design and three-dimensional modeling software. This is achieved by converting the geometric information inside and outside the vessel into a three-dimensional geometric model. Based on the vascular geometric model, a vascular mesh is generated for numerical simulation. The vascular model is discretized into small geometric units, such as a finite element mesh or a finite difference mesh. This can be accomplished using mesh generation software and algorithms. Based on the simulated blood flow data and the simulated blood shear stress data, a numerical simulation method is used to simulate the vascular flow. Computational fluid dynamics methods (such as the finite element method and the finite difference method) can be used to solve the fluid dynamics equations and simulate the blood flow within the coronary arteries. Based on the simulated blood flow data and the simulated blood shear stress data, appropriate boundary conditions and initial conditions are set. Then, the flow within the vessel is simulated using numerical methods for solving the fluid dynamics equations. This can be achieved using numerical simulation software and algorithms. Based on the numerical simulation results, a visualization of the three-dimensional coronary artery blood flow model is generated. This can be achieved by processing and presenting the simulation data using 3D visualization software and tools. The visualization results can provide an intuitive understanding and analysis of blood flow within the coronary arteries.

[0098] Preferably, step S33 includes the following steps:

[0099] Step S331: Analyze abnormal frames of the dynamic coronary angiography simulation video to obtain an abnormal lesion initial frame image and an abnormal lesion end frame image; confirm the lesion area based on the abnormal lesion initial frame image and the abnormal lesion end frame image to obtain lesion area position data;

[0100] Step S332: Classifying the lesion type of the lesion abnormal initial frame image and the lesion abnormal end frame image according to the lesion area position data to obtain lesion area type classification data; extracting lesion area quantitative features from the lesion area position data based on the lesion area type classification data to obtain lesion area vascular feature data;

[0101] Step S333: Analyze the blood flow situation in the lesion area using the coronary artery integrated model on the lesion area location data to obtain arterial blood flow characteristic data; perform severity analysis on the lesion area vascular characteristic data and arterial blood flow characteristic data to obtain coronary artery abnormality analysis data.

[0102] The present invention analyzes abnormal frames in dynamic coronary angiography simulation videos to obtain images of initial and final lesion abnormalities. These images provide information on the temporal and spatial changes of coronary lesions. Based on these images, the location data of the lesion region can be determined. This allows for accurate localization of the specific location of the coronary lesion. Based on the lesion region location data, the images of initial and final lesion abnormalities can be classified into lesion types. This facilitates classification of different types of coronary lesions, such as plaques and stenosis. Feature extraction of the lesion region classification data yields vascular feature data for the lesion region. These features can be used to describe quantitative information such as lesion morphology, size, and shape. Blood flow analysis of the lesion region location data using a coronary artery integrated model yields arterial blood flow feature data. This facilitates assessment of the impact of coronary lesions on hemodynamics. By comprehensively considering the lesion region vascular feature data and arterial blood flow feature data, the severity of the coronary abnormality can be analyzed. This helps physicians assess the extent of coronary lesions and their potential impact on patient health.

[0103] As an example of the present invention, refer to Figure 4 As shown, in this example, step S33 includes:

[0104] Step S331: Analyze abnormal frames of the dynamic coronary angiography simulation video to obtain an abnormal lesion initial frame image and an abnormal lesion end frame image; confirm the lesion area based on the abnormal lesion initial frame image and the abnormal lesion end frame image to obtain lesion area position data;

[0105] In an embodiment of the present invention, relevant videos of patients undergoing coronary angiography are obtained, which may be videos recorded by medical equipment or computer-generated videos simulating dynamic coronary angiography. Abnormal frames are detected in the dynamic coronary angiography simulation video. This may include using computer vision and image processing techniques, such as motion detection, edge detection, background modeling, etc., to detect abnormal frames in the video. Based on the abnormal frame detection results, the initial frame image and the ending frame image of the abnormal lesion are extracted from the dynamic coronary angiography simulation video. These frame images may show characteristics of coronary artery lesions, such as blood flow obstruction, plaque formation, etc. Based on the abnormal lesion frame images, image processing and analysis methods are used to confirm the lesion area. This may include using image segmentation algorithms, edge detection algorithms, etc. to determine the location of the lesion area. The location data of the lesion area is extracted from the lesion area confirmation results. This may include coordinate information, size information, shape information, etc. of the lesion area.

[0106] Step S332: Classifying the lesion type of the lesion abnormal initial frame image and the lesion abnormal end frame image according to the lesion area position data to obtain lesion area type classification data; extracting lesion area quantitative features from the lesion area position data based on the lesion area type classification data to obtain lesion area vascular feature data;

[0107] In embodiments of the present invention, lesion type classification is performed on lesion abnormality initial and end frame images using lesion location data. This can be based on expert knowledge or available lesion identification algorithms. For example, different lesion types, such as stenosis, plaque, and thrombus, can be classified by determining the morphological characteristics, density changes, and hemodynamic characteristics of the lesion area. Based on the lesion area type classification data, quantitative feature extraction is performed on the lesion area location data. This may include the following aspects: extracting the lesion area's external features, such as area, perimeter, and aspect ratio; extracting lesion intensity features, such as mean intensity and standard deviation, by analyzing the pixel grayscale values of the lesion area; extracting texture features of the lesion area using texture analysis methods, such as gray-level co-occurrence matrix and local binary pattern analysis; and extracting vascular flow characteristics by analyzing data such as blood flow velocity and blood pressure in the lesion area. Vascular feature data of the lesion area is extracted from the lesion area's quantitative features. This may include various feature data of the lesion area, such as shape, intensity, texture, and hemodynamic characteristics.

[0108] Step S333: Analyze the blood flow situation in the lesion area using the coronary artery integrated model on the lesion area location data to obtain arterial blood flow characteristic data; perform severity analysis on the lesion area vascular characteristic data and arterial blood flow characteristic data to obtain coronary artery abnormality analysis data.

[0109] In an embodiment of the present invention, an integrated coronary artery model is established that simulates the hemodynamic characteristics of the coronary arteries. The model can be implemented based on computational fluid dynamics (CFD) methods or other suitable numerical simulation techniques. The model needs to consider the following factors: constructing a vascular model based on vascular geometry data obtained from the location data of the lesion area. Simulating blood flow characteristics, such as flow velocity and pressure. Based on the established integrated coronary artery model, the location data of the lesion area is input into the model for simulation to obtain the blood flow conditions in the lesion area. The simulation results can be used to obtain arterial blood flow characteristic data, including blood flow velocity, pressure gradient, resistance, etc. The vascular characteristic data of the lesion area and the arterial blood flow characteristic data are combined to perform a severity analysis of the coronary abnormality, which involves assessing the extent of the lesion, the degree of vascular stenosis, and the magnitude of blood flow resistance based on the characteristics of the lesion area and the blood flow characteristics. The assessment can be performed based on existing diagnostic criteria for coronary artery disease or empirical rules.

[0110] Preferably, step S4 includes the following steps:

[0111] Step S41: reconstructing the coronary artery model of the coronary artery integrated model using the deep exploration data of the high abnormality part and the shallow exploration data of the low abnormality part to generate a coronary artery lesion model;

[0112] Step S42: Perform a virtual surgery simulation on the coronary artery lesion model to obtain a virtual surgery plan; use the virtual surgery plan to predict the effect of the coronary artery lesion model, thereby generating coronary artery effect prediction data; and construct a corresponding implementation plan template for the coronary artery lesion model based on the coronary artery effect prediction data.

[0113] The present invention reconstructs an integrated coronary artery model using deep exploration data from highly abnormal areas and shallow exploration data from less abnormal areas to generate a coronary artery lesion model. The deep exploration data is used to obtain structural information about the deeper layers of the coronary arteries, while the shallow exploration data is used to obtain structural information about the more superficial layers. By integrating these two data sets, a more accurate coronary artery model can be reconstructed. The coronary artery lesion model is then used to perform virtual surgical simulations. By setting different surgical plans, the effects of coronary artery treatment under different circumstances can be simulated. This can help doctors and researchers predict the therapeutic effects of different treatments on coronary artery lesions. Based on the virtual surgical plans and the coronary artery lesion model, coronary artery outcome predictions can be performed and coronary artery outcome prediction data can be generated. This data provides information on the potential therapeutic effects of different treatment plans. Based on the coronary artery outcome prediction data, corresponding implementation plan templates can be constructed. These templates can help doctors formulate treatment plans during actual surgeries and provide guidance and reference for achieving better therapeutic results.

[0114] In an embodiment of the present invention, deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas are collected. These data may come from medical imaging methods such as computed tomography (CT) or magnetic resonance imaging (MRI). The collected data is preprocessed, including image denoising and contrast enhancement, to improve data quality and clarity. Using techniques such as computer-aided reconstruction (CAD), the deep exploration data of highly abnormal areas and the shallow exploration data of less abnormal areas are integrated to construct an integrated coronary artery model. Based on the integrated coronary artery model, the coronary artery model is reconstructed to generate a coronary artery lesion model. This provides an accurate three-dimensional model that reflects the patient's coronary artery lesion condition. The coronary artery lesion model is used for virtual surgical simulation. Based on the specific case and treatment goals, different surgical plans are set, such as angioplasty or stenting. In the virtual surgical simulation, the procedures of different surgical plans are simulated, and the models are processed to simulate the impact and effect of the surgery on the coronary artery lesions. The effectiveness of the virtual surgical plans is evaluated, and the treatment outcomes of different surgical plans are predicted. This can be achieved by calculating hemodynamic parameters and lesion improvement in the model. Based on the virtual surgical plan and coronary artery outcome prediction data, a corresponding implementation plan template is constructed. This template will include information such as the specific surgical steps, required equipment and materials, and treatment goals to assist physicians in formulating and executing the treatment plan during the actual surgery.

[0115] The present invention provides the following advantages: by processing a patient's coronary artery scan images, a three-dimensional coronary artery spatial model can be constructed and vessel diameter analysis can be performed. This helps doctors understand the morphology, structure, and vessel diameter of the patient's coronary arteries. By processing the patient's X-ray angiography images, a three-dimensional coronary artery vascular model can be constructed and vessel wall thickness analysis can be performed. This helps doctors understand the vascular structure and wall thickness of the patient's coronary arteries. Based on the coronary artery wall thickness data and coronary artery diameter data, a three-dimensional coronary artery blood flow model can be constructed. This facilitates simulation and analysis of blood flow within the patient's coronary arteries. By integrating the three-dimensional coronary artery spatial model, the coronary artery vascular model, and the coronary artery blood flow model, an integrated coronary artery model is generated. By performing anomaly analysis on the integrated coronary artery model, deep exploration data for highly abnormal areas and shallow exploration data for less abnormal areas can be determined, thereby providing more accurate information on coronary artery lesions. Using the deep exploration data for highly abnormal areas and the shallow exploration data for less abnormal areas, the integrated coronary artery model is reconstructed to generate a coronary artery lesion model. The coronary artery lesion model can then be used to predict the effectiveness of virtual surgery, assessing the effectiveness of different treatment options and possible corrective surgery plans for coronary artery lesions. Therefore, the present invention improves the accuracy and reliability of coronary artery model reconstruction by combining a patient's coronary artery scan images and X-ray angiography images for multimodal modeling and performing coronary artery abnormality analysis.

[0116] The foregoing description is intended only to provide specific embodiments of the present invention, which will enable those skilled in the art to understand and implement the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not intended to be limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

Claims

1. A method for reconstructing a coronary artery model based on coronary angiography, characterized in that: The following steps are involved: Step S1: Acquire a coronary artery scan image of the patient; construct a three-dimensional coronary artery spatial structure on the patient's coronary artery image to generate a three-dimensional coronary artery spatial model; perform vascular diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vascular diameter data; Step S2: Acquire an X-ray angiography image of the patient; construct a coronary artery vascular structure based on the X-ray angiography image of the patient to generate a three-dimensional coronary artery vascular model; perform a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery vascular wall thickness data; Step S3: Constructing a coronary artery flow structure based on the coronary artery wall thickness data and the coronary artery diameter data to generate a three-dimensional coronary artery blood flow model; integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing coronary artery abnormality analysis on the coronary artery integrated model to generate deep exploration data of highly abnormal areas and shallow exploration data of less abnormal areas; Step S4: reconstruct the coronary artery model of the coronary artery integrated model using the deep exploration data of the highly abnormal part and the shallow exploration data of the low abnormal part to generate a coronary artery lesion model; predict the effect of virtual surgery on the coronary artery lesion model to construct a corresponding implementation plan template.

2. The method for reconstructing a coronary artery model based on coronary angiography according to claim 1, characterized in that: Step S1 includes the following steps: Step S11: obtaining a coronary artery scan image of the patient; Step S12: performing image denoising on the patient's coronary artery image to generate a coronary artery denoised image; performing image contrast enhancement on the coronary artery denoised image to generate a coronary artery enhanced image; Step S13: performing image vascular region segmentation on the coronary artery enhanced image to generate a coronary artery vascular region segmentation image; performing image binarization on the coronary artery vascular region segmentation image to generate a coronary artery vascular region binarization image; Step S14: constructing a three-dimensional coronary artery spatial structure based on the binarized image of the coronary artery region to generate a three-dimensional coronary artery spatial model; performing a vessel diameter analysis on the three-dimensional coronary artery spatial model to generate coronary artery vessel diameter data.

3. The method for reconstructing a coronary artery model based on coronary angiography according to claim 2, characterized in that: Step S14 includes the following steps: Step S141: extracting the centerline of the coronary artery region from the binary image to obtain the centerline of the coronary artery region; skeletonizing the coronary artery region from the binary image based on the centerline to generate a coronary artery skeleton image; Step S142: performing skeleton image pixel value change analysis on the coronary artery skeleton image to generate pixel change values in the direction of the vessel centerline; performing radial projection on the coronary artery skeleton image based on the pixel change values in the direction of the vessel centerline to generate radial projection data of the coronary artery main vessel; Step S143: confirming the coronary artery position based on the coronary artery skeleton image to obtain coronary artery position data; performing vascular tortuosity analysis based on the coronary artery position data and the coronary artery main vessel radial projection data to generate coronary artery curvature data; Step S144: Using the coronary artery curvature data, the coronary artery skeleton image is subjected to vascular stent detection to generate coronary artery branch point location data; based on the coronary artery branch point location data and the coronary main vessel radial projection data, a three-dimensional coronary artery spatial structure is constructed to generate a three-dimensional coronary artery spatial model; and the three-dimensional coronary artery spatial model is subjected to vascular diameter analysis to generate coronary artery diameter data.

4. The method for reconstructing a coronary artery model based on coronary angiography according to claim 1, characterized in that: Step S2 includes the following steps: Step S21: Acquire an X-ray angiography image of the patient; Step S22: performing image preprocessing on the patient X-ray angiography image to generate a standard patient X-ray angiography image, wherein the image preprocessing includes image denoising, image brightness enhancement, and image edge sharpening; Step S23: constructing the coronary artery vascular structure of the standard patient X-ray angiography image according to the coronary artery diameter data to generate a three-dimensional coronary artery vascular model; Step S24: performing a vascular wall thickness analysis on the three-dimensional coronary artery vascular model to generate coronary artery wall thickness data.

5. The method for reconstructing a coronary artery model based on coronary angiography according to claim 4, characterized in that: Step S23 includes the following steps: Step S231: hierarchically dividing the standard patient X-ray angiography image into angiographic images to obtain a surface angiography image and a deep angiography image; dividing the surface angiography image and the deep angiography image into overlapping regions to generate an angiography overlapping region image and an angiography non-overlapping region image; Step S232: performing initial vascular choroid connection on the non-overlapping region images of the angiography according to the coronary artery diameter data to obtain an initial vascular choroid image; Step S233: using the initial blood vessel venous image to predict the connection path of the overlapping region of the angiography overlapped region image, and generating blood vessel overlapped region connection path prediction data; Step S234: Based on the vascular overlapping area connection path prediction data and the coronary artery vascular curvature data, the vascular connection of the angiography overlapping area image is verified to generate vascular connection verification result data; the coronary artery vascular vascular structure is constructed for the angiography overlapping area image and the angiography non-overlapping area image using the vascular connection verification result data to generate a three-dimensional coronary artery vascular model.

6. The method for reconstructing a coronary artery model based on coronary angiography according to claim 5, characterized in that: Step S233 includes the following steps: Step S2331: using the initial vascular choroid image to segment the vascular connection ports of the overlapping region image of the angiography to obtain the vascular connection port data of the overlapping region; dividing the vascular connection port data of the overlapping region into a data set to generate a model training set and a model test set; Step S2332: performing model training on the model training set based on the support vector machine algorithm to generate a training model for the connection paths of the vascular overlapping regions; performing model testing on the training model for the connection paths of the vascular overlapping regions using the model test set to generate a prediction model for the connection paths of the vascular overlapping regions; Step S2333: importing the overlapping region blood vessel connection port data into the blood vessel overlapping region connection path prediction model to perform overlapping region blood vessel connection path prediction, thereby generating overlapping region blood vessel connection path prediction data.

7. The method for reconstructing a coronary artery model based on coronary angiography according to claim 1, characterized in that: Step S3 includes the following steps: Step S31: performing hemodynamic flow simulation based on the coronary artery wall thickness data and the coronary artery diameter data to generate coronary artery hemodynamic flow simulation data; constructing a coronary artery flow structure based on the coronary artery hemodynamic flow simulation data to generate a three-dimensional coronary artery blood flow model; Step S32: integrating the three-dimensional coronary artery spatial model, the three-dimensional coronary artery venous model, and the three-dimensional coronary artery blood flow model to generate a coronary artery integrated model; performing dynamic coronary angiography simulation on the coronary artery integrated model to generate a dynamic coronary angiography simulation video; Step S33: Analyzing the arterial blood flow condition of the dynamic coronary angiography simulation video to generate arterial blood flow characteristic data; performing coronary artery abnormality analysis on the coronary artery integrated model based on the arterial blood flow characteristic data to obtain coronary artery abnormality analysis data; Step S34: Compare the coronary artery abnormality analysis data with the preset standard abnormality threshold. When the coronary artery abnormality analysis data is greater than or equal to the preset standard abnormality threshold, a deep exploration of the coronary artery site of the patient is performed based on the coronary artery abnormality analysis data to obtain deep exploration data of the high abnormality site; when the coronary artery abnormality analysis data is less than the preset standard abnormality threshold, shallow exploration data of the low abnormality site is generated.

8. The method for reconstructing a coronary artery model based on coronary angiography according to claim 7, characterized in that: Step S31 The following steps are involved: Step S311: performing global external vascular feature screening on the three-dimensional coronary artery choroid model and the three-dimensional coronary artery spatial model according to the coronary artery wall thickness data and the coronary artery diameter data to obtain global external vascular feature data; Step S312: performing a virtual endoscopy based on the global external vessel length data, the global external vessel diameter data, and the global external vessel tortuosity data, thereby obtaining global internal vessel characteristic data; setting blood flow boundary conditions based on the global external vessel characteristic data and the global internal vessel characteristic data, thereby obtaining vascular blood flow boundary condition data; Step S313: Calculating the vascular flow velocity field using the vascular blood flow boundary condition data to generate blood flow velocity field data; analyzing the vascular pressure distribution based on the blood flow velocity field data and the vascular blood flow boundary condition data to generate coronary artery hemodynamic flow simulation data, wherein the coronary artery hemodynamic flow simulation data includes blood flow simulation data and blood flow shear stress simulation data; Step S314: constructing the coronary artery blood flow structure using the blood flow simulation data and the blood flow shear stress simulation data to generate a three-dimensional coronary artery blood flow model.

9. The method for reconstructing a coronary artery model based on coronary angiography according to claim 7, characterized in that: Step S33 includes the following steps: Step S331: Analyze abnormal frames of the dynamic coronary angiography simulation video to obtain an abnormal lesion initial frame image and an abnormal lesion end frame image; confirm the lesion area based on the abnormal lesion initial frame image and the abnormal lesion end frame image to obtain lesion area position data; Step S332: Classifying the lesion type of the lesion abnormal initial frame image and the lesion abnormal end frame image according to the lesion area position data to obtain lesion area type classification data; extracting lesion area quantitative features from the lesion area position data based on the lesion area type classification data to obtain lesion area vascular feature data; Step S333: Analyze the blood flow situation in the lesion area using the coronary artery integrated model on the lesion area location data to obtain arterial blood flow characteristic data; perform severity analysis on the lesion area vascular characteristic data and arterial blood flow characteristic data to obtain coronary artery abnormality analysis data.

10. The method for reconstructing a coronary artery model based on coronary angiography according to claim 1, characterized in that: Step S4 includes the following steps: Step S41: reconstructing the coronary artery model of the coronary artery integrated model using the deep exploration data of the high abnormality part and the shallow exploration data of the low abnormality part to generate a coronary artery lesion model; Step S42: Perform a virtual surgery simulation on the coronary artery lesion model to obtain a virtual surgery plan; use the virtual surgery plan to predict the effect of the coronary artery lesion model, thereby generating coronary artery effect prediction data; and construct a corresponding implementation plan template for the coronary artery lesion model based on the coronary artery effect prediction data.

Images