An automatic pre-processing method for constructing a fluid domain grid based on a blood vessel wall facet

Abstract

The application discloses an automatic preprocessing method for constructing a fluid domain grid based on a blood vessel wall facet, and comprises the following steps: a blood vessel internal imaging device generates a three-dimensional blood vessel wall facet file and converts the blood vessel wall facet file into an.obj format; a blood vessel wall facet file is recognized and processed, openings are recognized, and the openings are sequentially numbered; each recognized opening is repaired to form a watertight facet with color attributes; and a three-dimensional watertight blood fluid domain body grid is output. The method directly constructs a fluid domain body grid based on facet information, does not need to perform secondary processing such as surface fitting and smoothing on the blood vessel wall, maintains the accurate structure of the blood vessel lumen, and is beneficial to subsequent high-precision numerical calculation. Through marking of the repaired facet, boundary information is determined, manual analysis and boundary designation through an interactive interface are not needed, processing steps are simplified, and preprocessing time and operation complexity are reduced.

Description

Technical Field

[0001] This invention belongs to the field of diagnostic technology, specifically an automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets. Background Technology

[0002] Currently in my country, hemodynamic results are commonly used in the diagnosis and risk assessment of coronary artery disease. For example, simulations are used to calculate the fractional flow reserve (FFR) to assess whether coronary artery stenosis leads to myocardial ischemia, to evaluate the functional aspects of stenotic lesions, and to compensate for the limitations of traditional imaging, which only shows anatomical stenosis. This optimizes interventional treatment strategies, such as simulating the impact of different stent sizes and placement locations on blood flow to optimize stent selection. Numerical simulations validate device performance in virtual environments, accelerating development cycles and reducing experimental costs, thus promoting the development of new devices. Postoperative outcomes are assessed, such as predicting blood flow recovery after stent implantation and reducing the risk of restenosis. Research into pathological mechanisms is deepened, for example, simulating the relationship between low shear stress, blood flow turbulence, and endothelial cell dysfunction to reveal the biomechanical mechanisms of plaque formation; combined with coronary microcirculation resistance simulation, non-stenotic causes of myocardial ischemia are explored.

[0003] Coronary artery hemodynamic simulation (CFD) is a method that uses computer modeling and numerical simulation techniques to quantitatively analyze hemodynamic parameters (such as pressure, flow velocity, and shear force) within the coronary arteries. It has multiple important applications in medical research and clinical practice; hemodynamic results can assist in the diagnosis and risk assessment of coronary heart disease. The preprocessing step in CFD simulation is a crucial process that transforms medical imaging data into a fluid dynamics model suitable for numerical calculation, directly affecting the accuracy and reliability of the simulation results.

[0004] Traditional pretreatment steps such as Figure 1 As shown, the steps and difficulties of each step are as follows:

[0005] 1. Acquisition of medical imaging data:

[0006] Currently, the main coronary angiography techniques include CTA (providing three-dimensional anatomical structures with a resolution of approximately 0.5 mm), MRI (magnetic resonance imaging, suitable for patients without radiation requirements, but with lower resolution), and IVUS / OCT (intravascular ultrasound / optical coherence tomography, providing high-precision analysis of lumen and plaque morphology). Coronary vessels typically have complex and irregular structures; therefore, dual-modal IVUS-OCT three-dimensional imaging technology, which combines high penetration and high resolution, is beneficial for constructing high-precision vascular models, resulting in more reliable and accurate hemodynamic analysis of diseased vessels.

[0007] 2. Image Segmentation

[0008] The geometric structure of the coronary artery lumen is extracted to generate a three-dimensional surface model. Current imaging equipment can construct a three-dimensional vessel wall model, including branches and lumens, after image acquisition and use it for visualization.

[0009] 3. Geometric Repair and Simplification

[0010] The repair process mainly includes filling holes, smoothing surfaces (Laplacian smoothing or Taubin filtering), and surface fitting. Curvature features need to be preserved during repair, but simplification may lead to the loss of blood flow features, such as eddies in narrow areas. Currently, geometric processing mainly relies on third-party software, including Geomatic, Solidworks, SCDM, and MeshLab.

[0011] 4. Computational Domain Extraction

[0012] The repaired geometric model needs to meet the watertight condition. The region containing the geometric model is the fluid domain, which is also the solution domain for hemodynamic simulation. The geometric model needs to further specify the blood vessel inlet and outlet, as well as the blood vessel wall. This process is mainly carried out in the modeling software, where the user manually identifies and inputs the specified boundary information through an interactive interface.

[0013] 5. Meshing

[0014] Meshing the solution domain mainly consists of two parts: surface meshing and volume meshing. Mesh types include tetrahedral meshes, hybrid meshes, and hexahedral meshes. Volume meshing requires boundary layer meshing, local refinement in narrow regions, and overall mesh quality indicators to ensure the smooth progress of subsequent solutions. Currently, meshing techniques primarily rely on preprocessing software such as ANSYS ICEM, SimVascular, STAR-CCM+, and Hypermesh.

[0015] 6. Solving and Post-processing

[0016] Boundary conditions, material properties, and physical models are set according to the research scenario. Numerical calculations are performed using a CFD solver, and the simulation results are post-processed.

[0017] In summary, current coronary hemodynamic simulations place high demands on geometric model construction and mesh generation, and their automation levels are currently low. For example, hemodynamic simulations primarily rely on mature commercial software, such as modeling software SCDM, Solidworks, and Geomatic; and meshing software Hypermesh, ICEM, and Fluent Meshing. All of these software programs require data export, preprocessing, and migration for modeling, which is time-consuming, expensive, and has many limitations. Users need to perform model processing, mesh generation, and parameter settings step-by-step through an interactive interface, resulting in low automation and requiring a high level of user proficiency. These issues make it difficult for clinicians to operate directly, limiting the application of hemodynamic technology in the clinical field. Therefore, there is currently a lack of a preprocessing procedure for hemodynamic simulation that provides high-quality geometric and physical field inputs to the subsequent solver; and a solution is needed to address the current challenge of balancing computational efficiency with model accuracy. Summary of the Invention

[0018] To address the problems existing in the background technology, this invention provides an automated preprocessing method for constructing fluid domain meshes based on blood vessel wall facets. This method extracts fluid domains from complex blood vessel structures based on facet information, identifies inlets / outlets and walls, and constructs high-precision surface and volume meshes, thereby achieving automated preprocessing for hemodynamic simulation. The technical solution includes:

[0019] Intravascular imaging equipment generates three-dimensional vessel wall profile files and converts these files into .obj format;

[0020] The process involves identifying and processing the vessel wall facets in the vessel wall facet file, identifying openings, and sequentially numbering the openings.

[0021] Each identified opening is repaired to form a watertight facet with color attributes;

[0022] Output a three-dimensional watertight blood fluid domain volume mesh.

[0023] The intravascular imaging device is an OCT-IVUS dual-modal intravascular imaging device, which includes image groups of three modalities: IVUS single modality, OCT single modality, and OCT-IVUS dual modality. The intravascular imaging system automatically generates a vessel wall profile file, which includes geometric features.

[0024] The process of identifying and processing the vascular wall facets in the vascular wall facet file includes: processing the topological information of the vascular wall facets, identifying points, lines, and surfaces, and assigning the same initial color value to all identified points during the identification process;

[0025] The process of identifying openings includes: traversing the identified lines; if a line exists only in one triangular facet, then this line is an opening boundary line; analyzing the connection information of all boundary lines; if a group of boundary lines are connected end to end to form a closed loop, then this boundary line constitutes an opening; in the process of determining the boundary line: if a line is shared by two triangular facets, then this line is an internal line; if a line is used only by a single triangular facet, then this line is located on the boundary and is marked as a boundary line.

[0026] The process of sequentially numbering the openings includes: sorting the openings according to the coordinates of the center point of the opening and numbering them sequentially as opening numbers.

[0027] Each opening is repaired using a mesh repair tool. The repair process for each opening is as follows: points and triangular facets are generated within the closed loop area to fill the entire opening area, serving as tokens for each opening to form the opening boundary for subsequent blood flow simulation.

[0028] The mesh size for generating triangular facets is determined by the size of the line segments on the loop, to ensure that the mesh density of the generated facets is consistent with the overall mesh density.

[0029] The assigned color value is determined by the opening number, and points on the same opening have the same color value.

[0030] The process of directly performing volume mesh generation on watertight facets with color attributes is as follows: based on the repaired facets and the color values ​​of the extracted boundary information, the watertight conditions have been met, and volume mesh generation is performed.

[0031] The output 3D watertight blood fluid domain volume mesh is interfaced with the solver, and blood flow simulation is performed.

[0032] The output 3D watertight blood fluid domain volume mesh is displayed directly on the monitor.

[0033] After numbering the openings sequentially, the following steps are taken: The opening locations are named according to their types: Based on the opening coordinates, the opening type is determined. If the opening is at the rightmost end of the entire model, it is an "entrance"; if the opening is at the leftmost end of the entire model, it is an "exit"; and all other openings are "branches".

[0034] When outputting the three-dimensional watertight blood fluid domain volume mesh, the name of each opening location is simultaneously displayed on the monitor.

[0035] The beneficial effects of this invention are as follows:

[0036] 1. This method directly constructs a fluid domain volume mesh based on facet information, eliminating the need for secondary processing such as surface fitting and smoothing of the vessel wall. This preserves the accurate structure of the vessel lumen, which is beneficial for subsequent high-precision numerical calculations. By marking the repaired facets, boundary information is determined, eliminating the need for manual analysis and boundary specification through an interactive interface. This simplifies the processing steps, reduces preprocessing time and operational complexity, and builds a technical bridge between intravascular imaging technology and hemodynamic simulation.

[0037] 2. It has higher adaptability and robustness; the opening identified based on intravascular imaging may be uneven (e.g., Figure 7 As shown in the diagram, the preprocessing process often requires the operator to manually segment and repair the opening surface before subsequent operations such as opening numbering and blood flow simulation can be performed. However, this invention identifies boundaries through topological structure and automatically generates and numbers the openings. Therefore, unlike traditional cutting methods that require users to manually repair, cut openings, and determine a flat opening surface within a commercial software's graphical interface, this invention reduces the loss of fine geometric information, simplifies preprocessing steps, reduces preprocessing time and operational complexity, significantly lowers operational difficulty, and improves user comfort.

[0038] 3. It eliminates the need for pre-processing software, eliminating the need for secondary processing operations such as re-meshing, smoothing, fine-tuning, opening selection, and boundary specification. No interactive interface is required during the repair process, thus achieving automated pre-processing for hemodynamic simulation. Attached Figure Description

[0039] Figure 1 This is a schematic diagram of the preprocessing steps in the prior art of the automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets, as described in this invention.

[0040] Figure 2 This is a schematic diagram of the process in an embodiment of the present invention.

[0041] Figure 3 This is a schematic diagram of the results of identifying points, lines, and surfaces in step 201 of this embodiment of the invention.

[0042] Figure 4 This is a schematic diagram of the generation results of steps two and three in an embodiment of the present invention.

[0043] Figure 5 This is a schematic diagram of the surface mesh division result in an embodiment of the present invention.

[0044] Figure 6 This is a schematic diagram of the volume mesh division result in an embodiment of the present invention.

[0045] Figure 7 This is a three-dimensional vascular geometric model obtained from an image by an intravascular imaging device in an embodiment of the present invention. Detailed Implementation

[0046] The present invention will be further described in detail below with reference to the accompanying drawings.

[0047] like Figures 2-6 The embodiment of the present invention shown includes:

[0048] Step 1: The intravascular imaging device generates a 3D vessel wall profile file and converts it to .obj format. Specifically, it uses an OCT-IVUS dual-modal intravascular imaging device, which includes image groups of three modalities: IVUS monomodal, OCT monomodal, and OCT-IVUS dual-modal. The intravascular imaging system automatically generates the vessel wall profile file, which includes geometric features such as branches and lumen shapes.

[0049] Step Two: Based on the topological structure of the vessel wall facets (based on the points, lines, and surfaces contained in the vessel wall facet file), utilizing all the triangular structural units that make up the vessel wall facets in the file, specifically, the points, lines, and surfaces contained in the vessel wall facet file are identified and processed differently to identify openings and boundaries, thereby marking the openings, sequentially numbering the openings, and determining their corresponding opening types; in this process, there is no need to manually specify the inlet and outlet through an interactive interface, achieving automated boundary recognition; specifically:

[0050] Step 201: Process the topological information of the vessel wall facets, identify points, lines, and surfaces. During the identification process, assign the same initial color value to all identified points. Points with initial color values ​​are all of type "vessel wall". Since the facet file first determines n points (nodes) on the vessel wall and the three-dimensional coordinates and number of each point; a line can be determined by the number of two points, and m lines can be constructed from the n points of the vessel wall; a triangular facet can be determined by three lines connected end to end. This is the smallest unit of the vessel wall facet. The vessel wall is composed of p three-part facets.

[0051] Step 202: Identify Openings: Traverse the identified lines. If a line exists only in one triangular facet, it is an opening. Lines on the boundary of an opening are boundary lines. Analyze the connection information of all boundary lines. If a group of boundary lines are connected end to end to form a closed loop, they constitute an opening. In the process of determining the boundary lines: if a line is shared by two triangular facets, it is an internal line; if a line is used only by a single triangular facet, it is located on the boundary and is marked as a boundary line.

[0052] Step 203: Number the openings sequentially: Sort the openings according to the coordinates of the center point of the opening and number them sequentially as opening numbers;

[0053] Step 204: Name the opening locations according to each opening type; determine the opening type based on the opening coordinates. If the opening is at the rightmost end of the entire model (Z coordinate is at its minimum value), then this opening is called "inlet"; if the opening is at the leftmost end of the entire model (Z coordinate is at its maximum value), then this opening is called "outlet"; all other openings are "branch".

[0054] Step 3: Repair each identified opening; during the opening repair process, generate mesh points and triangular facets within the opening, and assign the same color value to the mesh points according to the opening number (one of the built-in attributes of the .obj format), thus forming a watertight facet with color attributes; since the vessel wall facet only contains the surface structure of the vessel wall, and also has inlets and outlets such as branches, the extraction of the fluid domain first requires repairing the identified boundaries. In this step, by adding color attributes to the newly added points (nodes) in the .obj format geometry file, we ensure that the boundary information is retained while the geometry is modified, thus achieving automated processing of geometric model repair and boundary recognition;

[0055] The mesh repair tool is used to repair each opening. The repair process for each opening is as follows: points and triangular facets are generated within the closed loop area to fill the entire opening area, serving as tokens for each opening to form the opening boundary and connect to the subsequent blood flow simulation.

[0056] The mesh size for generating triangular facets is determined by the size of the line segments on the loop, thus ensuring that the mesh density for generating triangular facets is consistent with the overall mesh density.

[0057] The assigned color value is determined based on the opening number, meaning that points on the same opening have the same color value. The color value is an RGB value, and each RGB value contains three numbers, for example, red is (255, 0, 0). In this embodiment, the color value assignment process is as follows: the initial color value (the color value of the point on the blood vessel wall) is set to (0, 0, 0). If there are a total of 5 openings (inlet, outlet, and branch), then 5 RGB values ​​are generated: (51, 51, 51), (102, 102, 102), (153, 153, 153), (204, 204, 204), and (255, 255, 255). These five RGB values ​​are assigned to openings 1, 2, 3, 4, and 5 in sequence. The color values ​​only need to be distinguishable, clearly indicating the correspondence between the color value and the opening number.

[0058] Step 4: Directly perform volume mesh generation on the watertight facets with color attributes; since the facets repaired in Step 3 already meet the watertight condition and have color values ​​from which boundary information can be extracted; therefore, the model formed after the repaired opening is directly used as a surface mesh for volume mesh generation; thus eliminating the steps of smooth surface fitting and re-meshing the surface mesh, preserving the geometric features of the blood vessels, and realizing the automatic generation of the volume mesh.

[0059] The watertight facets formed after repairing the opening contain opening information and color values, and are directly used as face meshes for the next step of volume mesh generation.

[0060] Step 5: Output the volume mesh of the 3D watertight blood fluid domain for subsequent blood flow simulation or direct display on the screen: The generation of the volume mesh means that the preprocessing step of blood flow simulation is completed. The blood vessel wall facets in the watertight facets are in a watertight state, that is, the facets have no gaps and the interior is a fluid domain; this allows the output volume mesh to be interfaced with different types of solvers for subsequent blood flow simulation (coronary artery hemodynamic simulation).

[0061] Because the openings are colored, even operators unfamiliar with commercial software (such as clinicians or nurses) can distinguish between them. If the names are displayed, the type of opening can be identified directly.

Claims

1. An automated preprocessing method for constructing fluid domain meshes based on blood vessel wall facets, characterized in that, include: Intravascular imaging equipment generates three-dimensional vessel wall profile files and converts these files into .obj format; The process involves identifying and processing the vascular wall facets in the vascular wall facet file, identifying openings, and numbering the openings sequentially. The process of numbering the openings sequentially includes: sorting the openings according to the coordinates of the center point of the opening and numbering them sequentially as opening numbers. Each identified opening is repaired to form a watertight surface with color attributes. The process of directly dividing the watertight surface with color attributes into a volume mesh is as follows: the watertight conditions are met, and the repaired surface and the color value of the extracted boundary information are used to divide the volume mesh. Output a three-dimensional watertight blood fluid domain mesh; The process of identifying and processing the vascular wall facets in the vascular wall facet file includes: processing the topological information of the vascular wall facets, identifying points, lines, and surfaces, and assigning the same initial color value to all identified points during the identification process; The process of identifying openings includes: traversing the identified lines; if a line exists only in one triangular facet, then this line is an opening boundary line; analyzing the connection information of all boundary lines; if a group of boundary lines are connected end to end to form a closed loop, then this boundary line constitutes an opening; in the process of determining the boundary line: if a line is shared by two triangular facets, then this line is an internal line; if a line is used only by a single triangular facet, then this line is located on the boundary and is marked as a boundary line. Each opening is repaired using a mesh repair tool. The repair process for each opening is as follows: points and triangular facets are generated within the closed loop area to fill the entire opening area, serving as tokens for each opening to form the opening boundary for subsequent blood flow simulation. The mesh size for generating triangular facets is determined by the size of the line segments on the loop to ensure that the mesh density of the generated facets is consistent with the overall mesh density; After numbering the openings sequentially, the following steps are taken: The opening locations are named according to their types: Based on the opening coordinates, the opening type is determined. If the opening is at the rightmost end of the entire model, it is called an "entrance"; if the opening is at the leftmost end of the entire model, it is called an "exit"; and all other openings are "branches".

2. The automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets according to claim 1, characterized in that, The intravascular imaging device is an OCT-IVUS dual-modal intravascular imaging device, which includes image groups of three modalities: IVUS single modality, OCT single modality, and OCT-IVUS dual modality. The intravascular imaging system automatically generates a vessel wall profile file, which includes geometric features.

3. The automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets according to claim 1, characterized in that, The assigned color value is determined by the opening number, and points on the same opening have the same color value.

4. The automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets according to claim 1, characterized in that, The output 3D watertight blood fluid domain volume mesh is interfaced with the solver, and blood flow simulation is performed.

5. The automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets according to claim 1, characterized in that, The output 3D watertight blood fluid domain volume mesh is displayed directly on the monitor.

6. The automated preprocessing method for constructing a fluid domain mesh based on blood vessel wall facets according to claim 1, characterized in that, When outputting the three-dimensional watertight blood fluid domain volume mesh, the name of each opening location is simultaneously displayed on the monitor.

Images