Vessel centerline-based stenosis lesion three-dimensional reconstruction method and system

Abstract

The application discloses a stenosis lesion three-dimensional reconstruction method and system based on a blood vessel center line, wherein the method comprises the following steps: acquiring blood vessel center line data, a radius sequence corresponding to the center line and medical image data; determining a minimum radius value and a corresponding index position according to the radius sequence, and performing filtering processing on the radius sequence to obtain a smooth radius sequence; performing regression correction processing according to the minimum radius value and the smooth radius sequence to obtain a corrected radius sequence; acquiring a target region mask according to the medical image data, determining a center line point to be removed and a clipping flag according to the spatial position relationship between the blood vessel center line data and the target region mask; constructing a blood vessel mask in a voxel space according to the corrected radius sequence and the blood vessel center line data after removing the center line point to be removed; generating a three-dimensional grid model according to the blood vessel mask, and performing end face clipping processing on the three-dimensional grid model according to the clipping flag. The application forms a blood vessel three-dimensional reconstruction result with continuous structure and clear boundary.

Description

Technical Field

[0001] This application relates to the fields of medical image processing and computer-aided diagnosis technology, specifically to a three-dimensional reconstruction method for stenotic lesions based on the vascular centerline, a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline, an electronic device, and a readable storage medium. Background Technology

[0002] With the development of medical imaging technology, three-dimensional reconstruction of vascular structures based on tomographic image data has become an important technical means for clinical analysis of vascular morphology and assessment of vascular abnormalities. In related applications, vascular regions in medical images are typically segmented and combined with vascular centerline data to construct a spatial structural model of the blood vessels. However, due to the differences in inter-slice resolution in medical imaging data and the inherent complexity and curvature of vascular structures, direct three-dimensional reconstruction based on the original data can easily result in discontinuous structures on the surface of the vascular model. Furthermore, structural distortions can easily occur in tortuous and truncated areas of the blood vessels, thus affecting the stability of the overall vascular structural representation.

[0003] Furthermore, in vascular branching regions or vascular origin regions, there is often a spatial adjacency between vascular structures and adjacent tissues. Existing modeling methods, when determining the boundaries of vascular structures, easily include non-target tissues in the model or result in irregular morphology of the vascular end structures, affecting the integrity of the vascular model. Simultaneously, during vascular modeling, areas with significant changes in local vascular geometric parameters are easily affected by data processing, causing deviations between the vascular structure representation and the original image structure. Summary of the Invention

[0004] The purpose of this application is to provide a three-dimensional reconstruction method for stenotic lesions based on the vascular centerline, a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline, an electronic device, and a readable storage medium, which can solve the problems in the prior art of insufficient continuity of vascular structure, deviation in local structural expression, and difficulty in accurately determining the boundary between blood vessels and adjacent tissues in the process of 3D vascular modeling based on medical images.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows: In a first aspect, embodiments of this application provide a method for three-dimensional reconstruction of stenotic lesions based on the vascular centerline, the method comprising: Acquire vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data; The minimum radius and its corresponding index position are determined based on the radius sequence, and the radius sequence is filtered to obtain a smooth radius sequence. Based on the minimum radius value and the smoothed radius sequence, a regression correction process is performed to obtain the corrected radius sequence; The target region mask is obtained based on the medical image data, and the centerline point to be removed and the clipping marker are determined based on the spatial relationship between the blood vessel centerline data and the target region mask. Based on the corrected radius sequence and the vessel centerline data after removing the centerline points to be removed, a vessel mask is constructed in voxel space. A three-dimensional mesh model is generated based on the blood vessel mask, and the end face of the three-dimensional mesh model is clipped according to the clipping flag.

[0006] Optionally, the step of filtering the radius sequence to obtain a smooth radius sequence includes: Obtain the maximum and minimum radii of the radius sequence; The filtering parameters are determined based on the difference between the maximum radius and the minimum radius; The radius sequence is processed sequentially by median filtering and Gaussian filtering to obtain the smoothed radius sequence.

[0007] Optionally, the step of performing regression correction processing based on the minimum radius and the smoothed radius sequence to obtain the corrected radius sequence includes: Calculate the difference between the radius value of the smoothed radius sequence at the index position and the minimum radius value; When the difference meets the preset conditions, a decay correction function centered on the index position is constructed; The smooth radius sequence is corrected according to the attenuation correction function to obtain the corrected radius sequence.

[0008] Optionally, determining the centerline point to be removed and the clipping marker based on the spatial relationship between the blood vessel centerline data and the target area mask includes: Map the physical coordinates of the centerline points in the blood vessel centerline data to voxel coordinates; Determine whether each centerline point is located within the target region mask based on the voxel coordinates; The proportion of centerline points located within the target area mask is statistically analyzed, and the centerline points to be removed are determined based on the comparison between the proportion and a preset proportion threshold. The clipping flag is set based on the removal status of the centerline point corresponding to the starting index position in the vascular centerline data.

[0009] Optionally, constructing a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed includes: The blood vessel centerline data are subjected to interpolation sampling processing; Based on the voxel spacing parameters of the medical image data, calculate the physical distance of each voxel to the centerline point in the vascular centerline data; Determine the nearest centerline projection point for each voxel in the vascular centerline data, and obtain the correction radius corresponding to the nearest centerline projection point; When the physical distance is less than the corrected radius, the voxel is marked as a vascular region voxel, and the vascular mask is constructed based on all the marked vascular region voxels.

[0010] Optionally, after constructing the vascular mask based on all marked vascular region voxels, the method further includes: Obtain the narrowing rate value corresponding to the nearest centerline projection point; The stenosis rate value is converted into an integer label value according to a preset mapping relationship and written into the vascular region voxel.

[0011] Optionally, the step of generating a three-dimensional mesh model based on the vascular mask and performing end-face clipping processing on the three-dimensional mesh model according to the clipping flag includes: The blood vessel mask is blurred, and a triangular mesh model is generated based on the isosurface extraction algorithm; Based on the tangent vector of the centerline point at the endpoint in the blood vessel centerline data after removing the centerline point to be removed, a clipping plane is constructed, and the clipping plane is used to perform geometric segmentation on the triangular mesh model to obtain at least two independent sub-mesh models. Calculate the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before performing the geometric segmentation process, and retain the sub-mesh model with the smaller distance.

[0012] Secondly, embodiments of this application provide a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline, the system comprising: The data acquisition module is used to acquire vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data; A smooth radius sequence determination module is used to determine the minimum radius value and its corresponding index position based on the radius sequence, and to filter the radius sequence to obtain a smooth radius sequence; The modified radius sequence determination module is used to perform regression correction processing based on the minimum radius value and the smoothed radius sequence to obtain the modified radius sequence; The line point marker determination module is used to obtain a target area mask based on the medical image data, and determine the center line points to be removed and the clipping markers based on the spatial positional relationship between the blood vessel centerline data and the target area mask. A vascular mask construction module is used to construct a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed. The trimming module is used to generate a three-dimensional mesh model based on the blood vessel mask, and to perform end-face trimming processing on the three-dimensional mesh model according to the trimming flag.

[0013] Optionally, the smooth radius sequence determination module includes: A radius acquisition module is used to acquire the maximum and minimum radii of the radius sequence. A filter parameter determination module is used to determine filter parameters based on the difference between the maximum radius and the minimum radius; The filtering module is used to sequentially perform median filtering and Gaussian filtering on the radius sequence to obtain the smoothed radius sequence.

[0014] Optionally, the corrected radius sequence determination module includes: The difference calculation module is used to calculate the difference between the radius value of the smoothed radius sequence at the index position and the minimum radius value; The function construction module is used to construct a decay correction function centered on the index position when the difference meets a preset condition; The smooth radius correction module is used to correct the smooth radius sequence according to the decay correction function to obtain the corrected radius sequence.

[0015] Optionally, the line dot flag determination module includes: The coordinate mapping module is used to map the physical coordinates of the centerline points in the blood vessel centerline data to voxel coordinates. The coordinate determination module is used to determine whether each centerline point is located within the target area mask based on the voxel coordinates. The centerline point determination module is used to statistically analyze the proportion of centerline points located within the target area mask, and determine the centerline points to be removed based on the comparison result between the proportion and a preset proportion threshold. The flag setting module is used to set the clipping flag based on the removal status of the centerline point corresponding to the starting index position in the vascular centerline data.

[0016] Optionally, the vascular mask construction module includes: An interpolation sampling module is used to perform interpolation sampling processing on the blood vessel centerline data; The physical distance calculation module is used to calculate the physical distance from each voxel to the centerline point in the blood vessel centerline data based on the voxel spacing parameter of the medical image data. The correction radius determination module is used to determine the nearest centerline projection point of each voxel in the blood vessel centerline data, and obtain the correction radius corresponding to the nearest centerline projection point; A vascular region voxel marking module is used to mark the voxel as a vascular region voxel when the physical distance is less than the correction radius, and to construct the vascular mask based on all the marked vascular region voxels.

[0017] Optionally, the system further includes: The stenosis rate acquisition module is used to acquire the stenosis rate value corresponding to the nearest centerline projection point after the vascular region voxel marking module constructs the vascular mask based on all marked vascular region voxels. The numerical conversion module is used to convert the stenosis rate value into an integer label value according to a preset mapping relationship and write it into the vascular region voxel.

[0018] Optionally, the cropping processing module includes: A triangular mesh model generation module is used to blur the blood vessel mask and generate a triangular mesh model based on an isosurface extraction algorithm. The sub-mesh model determination module is used to construct a clipping plane based on the tangent vector of the centerline point at the endpoint in the blood vessel centerline data after removing the centerline point to be removed, and to use the clipping plane to perform geometric segmentation processing on the triangular mesh model to obtain at least two independent sub-mesh models. The sub-mesh model retention module is used to calculate the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before the geometric segmentation process is performed, and to retain the sub-mesh model with the smaller distance.

[0019] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0020] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0021] In this embodiment, by acquiring the blood vessel centerline data and its corresponding radius sequence, and filtering and regressing the radius sequence under the constraint of minimum radius value, the blood vessel centerline points are screened in conjunction with the target region mask, and a blood vessel mask is constructed based on the corrected radius sequence and the screened blood vessel centerline data. Furthermore, the end face of the three-dimensional mesh model is processed according to the clipping flag, so that the constructed blood vessel model can maintain the continuous expression of the blood vessel spatial structure, while taking into account the determination of the blood vessel boundary range and the regular expression of the end structure, thereby making the generated three-dimensional blood vessel structure consistent with the blood vessel morphology in medical images. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the steps of a three-dimensional reconstruction method for stenotic lesions based on the vascular centerline according to an embodiment of this application. Figure 2 This is a schematic diagram illustrating the principle of a high-fidelity three-dimensional reconstruction and visualization scheme for stenotic lesions based on the vascular centerline according to an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline according to an embodiment of this application; Figure 4 This is a schematic diagram of the hardware structure of an electronic device according to an embodiment of this application. Detailed Implementation

[0023] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0024] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0025] This application provides a three-dimensional reconstruction scheme for stenotic lesions based on the vascular centerline. By acquiring vascular centerline data and its corresponding radius sequence, the radius sequence is corrected under constraints. Combined with the spatial information of the target area in medical imaging data, the vascular centerline is filtered. Based on this, a vascular mask is constructed and a corresponding three-dimensional mesh model is generated. At the same time, the model end face is processed according to the clipping marker, thereby forming a vascular three-dimensional reconstruction result with continuous structure and clear boundaries, which is suitable for vascular structure modeling scenarios under medical imaging conditions.

[0026] The following description, in conjunction with the accompanying drawings, details a three-dimensional reconstruction scheme for stenotic lesions based on the vascular centerline provided in this application, through specific embodiments and application scenarios.

[0027] Reference Figure 1 The diagram illustrates a flowchart of a three-dimensional reconstruction method for stenotic lesions based on the vascular centerline according to an embodiment of this application.

[0028] Step 101: Obtain the vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data.

[0029] First, vascular centerline data is acquired. This data describes the spatial orientation and topology of blood vessels in three-dimensional space. It typically consists of multiple centerline points arranged sequentially along the vessel path, each containing spatial coordinate information. The vascular centerline data reflects the relationship between the main trunk and branches of the vessel and serves as a reference path for subsequent vascular mask construction and model generation. Simultaneously, a radius sequence corresponding to the centerline is acquired. This radius sequence describes the scale information of the vessel cross-section at each centerline point and establishes a one-to-one correspondence with the vascular centerline data, ensuring continuous representation of the vascular geometry along the centerline direction. The radius sequence is usually stored according to the order of the centerline points in the vascular centerline data for unified processing of the radius distribution later. In addition, medical imaging data is acquired. This data provides voxel spatial information and tissue distribution information of the vascular structure, providing fundamental data support for subsequent target area mask acquisition, spatial positional relationship determination, and vascular mask construction. The medical imaging data also includes voxel spacing parameters, used for spatial distance calculation and structural localization in subsequent steps.

[0030] Step 102: Determine the minimum radius and its corresponding index position based on the radius sequence, and filter the radius sequence to obtain a smooth radius sequence.

[0031] Specifically, the acquired radius sequence is first traversed to determine the minimum radius and record its corresponding index position. This index position indicates the positional relationship of the minimum radius within the radius sequence, providing a locational basis for subsequent radius correction. After determining the minimum radius and its corresponding index position, the radius sequence is filtered to obtain a smoothed radius sequence. Filtering smooths local variations in the radius sequence, making the radius variation along the centerline more continuous, thus providing stable geometric scale information for subsequent construction of the vascular spatial structure. During filtering, the overall variation amplitude of the radius sequence can be considered to determine the filtering parameters, allowing the filtering process to adapt to the radius variation characteristics of different vascular structures.

[0032] Step 103: Perform regression correction processing based on the minimum radius and the smoothed radius sequence to obtain the corrected radius sequence.

[0033] First, the difference between the radius value at the index position and the minimum radius value of the smoothed radius sequence is calculated. Based on this difference, the smoothed radius sequence is locally corrected so that it remains consistent with or close to the minimum radius value at the index position, thus forming a corrected radius sequence. Regression correction maintains the overall continuous variation characteristics of the radius sequence while ensuring that the radius sequence reflects the original radius distribution characteristics at key positions. The corrected radius sequence is still arranged according to the order of the centerline points in the vascular centerline data and maintains its correspondence with the vascular centerline data.

[0034] Step 104: Obtain the target area mask based on medical imaging data, and determine the centerline point to be removed and the clipping marker position based on the spatial relationship between the vascular centerline data and the target area mask.

[0035] Specifically, the target region mask is used to represent the spatial extent of a specified tissue region in the medical imaging data. Subsequently, the spatial coordinates of the centerline points in the vascular centerline data are mapped to the voxel space corresponding to the medical imaging data. Based on the spatial relationship between the centerline points and the target region mask, it is determined whether each centerline point is located within the region represented by the target region mask. After completing the spatial position determination, centerline points to be removed are determined based on the distribution of centerline points, used to limit the centerline range of subsequent vascular structure modeling. Simultaneously, based on the positional relationship of the centerline points to be removed in the vascular centerline data, clipping flags are determined. These clipping flags are used to indicate whether end-face clipping processing should be performed on the subsequent 3D mesh model.

[0036] Step 105: Construct a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed.

[0037] Specifically, after obtaining the corrected radius sequence and determining the centerline points to be removed, the vessel centerline data is updated to obtain vessel centerline data after removing the centerline points to be removed. Subsequently, spatial distances are calculated in voxel space based on the vessel centerline data, and the spatial distance relationship between each voxel in the voxel space and the centerline point in the vessel centerline data is determined. Simultaneously, the radius information of the corresponding centerline point in the corrected radius sequence is obtained, and the voxel is used to determine the vessel region based on the distance relationship between the voxel and the centerline point and the corresponding radius information. When a voxel meets a preset distance relationship, it is marked as a vessel region voxel, and a vessel mask is constructed based on all vessel region voxels. The vessel mask is used to characterize the spatial distribution range of vessels in voxel space, providing a structural basis for subsequent 3D mesh model generation.

[0038] The aforementioned preset distance relationship is used to define whether a voxel belongs to the vascular spatial range. For example, in one implementation, when the physical distance from a voxel to the nearest centerline projection point is less than or equal to the correction radius corresponding to that projection point, the voxel is marked as a vascular region voxel. In another implementation, a fixed distance compensation value is introduced based on the correction radius, and the preset distance is set as the correction radius plus a preset offset. In yet another implementation, the distance threshold is adjusted proportionally according to the voxel spacing parameter in the medical image data to ensure that the preset distance is consistent with the voxel spatial resolution. By setting the aforementioned preset distance relationship, the determination of vascular region voxels maintains a correspondence with the vascular centerline data and the correction radius sequence.

[0039] Step 106: Generate a three-dimensional mesh model based on the blood vessel mask, and perform end-face clipping processing on the three-dimensional mesh model according to the clipping markers.

[0040] Specifically, firstly, a 3D mesh generation process is performed based on the vascular mask, converting the vascular region within the mask into a triangular mesh model, thus representing the spatial structure of the blood vessels in mesh form. Subsequently, end-face clipping is performed on the 3D mesh model according to clipping flags. During end-face clipping, clipping conditions are constructed based on the spatial orientation information at the endpoints of the vascular centerline data, and geometric segmentation is performed on the 3D mesh model to adjust the end-structure morphology. The clipping process limits the end-space range of the vascular model, ensuring that the generated 3D mesh model maintains consistency with the effective structural range corresponding to the vascular centerline data.

[0041] In this embodiment, by acquiring the blood vessel centerline data and its corresponding radius sequence, and filtering and regressing the radius sequence under the constraint of minimum radius value, the blood vessel centerline points are screened in conjunction with the target region mask, and a blood vessel mask is constructed based on the corrected radius sequence and the screened blood vessel centerline data. Furthermore, the end face of the three-dimensional mesh model is processed according to the clipping flag, so that the constructed blood vessel model can maintain the continuous expression of the blood vessel spatial structure, while taking into account the determination of the blood vessel boundary range and the regular expression of the end structure, thereby making the generated three-dimensional blood vessel structure consistent with the blood vessel morphology in medical images.

[0042] In one exemplary embodiment of this application, one way to filter the radius sequence to obtain a smooth radius sequence is as follows: obtain the maximum radius and minimum radius of the radius sequence; determine the filtering parameters based on the difference between the maximum radius and the minimum radius; and perform median filtering and Gaussian filtering on the radius sequence in sequence to obtain a smooth radius sequence.

[0043] In this embodiment, to stabilize the variation trend of the radius sequence along the vessel centerline, a comprehensive statistical analysis is first performed on the radius sequence to obtain the maximum and minimum radii. The maximum and minimum radii characterize the scale variation range of the vessel along the centerline and provide a basis for determining subsequent filtering parameters. After obtaining the maximum and minimum radii, filtering parameters are determined based on the difference between them. These parameters limit the window size and smoothness during filtering, enabling the filtering process to adapt to the radius variation characteristics of different vessel structures. After determining the filtering parameters, median filtering and Gaussian filtering are performed sequentially on the radius sequence. Median filtering is mainly used to adjust discrete outliers in the radius sequence. By sorting neighboring radius values ​​and selecting the median value to replace the original radius value, local abrupt changes in the radius sequence are mitigated. Gaussian filtering is then performed, continuously weighting the radius sequence along the centerline to create a continuous transition in radius variation. The combination of median filtering and Gaussian filtering balances local anomaly adjustment with overall smoothing. After the above filtering process, a smoothed radius sequence is obtained. The smoothed radius sequence still maintains a one-to-one correspondence with the blood vessel centerline data and provides basic data for subsequent regression correction processing.

[0044] This embodiment determines the filtering parameters by combining the maximum and minimum radii, and performs median filtering and Gaussian filtering in sequence to make the radius sequence change more continuously along the direction of the blood vessel centerline. At the same time, it adjusts the local abnormal changes in the radius sequence, providing a stable radius distribution basis for subsequent radius correction processing.

[0045] In one exemplary embodiment of this application, the method of performing regression correction processing based on the minimum radius value and the smoothed radius sequence to obtain the corrected radius sequence is as follows: calculate the difference between the radius value of the smoothed radius sequence at the index position and the minimum radius value; when the difference meets the preset condition, construct an attenuation correction function centered at the index position; and correct the smoothed radius sequence according to the attenuation correction function to obtain the corrected radius sequence.

[0046] In this embodiment, firstly, the difference between the radius value at the index position and the minimum radius value of the smoothed radius sequence is calculated to reflect the degree of change in the radius value at that position caused by the filtering process. When the difference meets a preset condition, an attenuation correction function is constructed centered on the index position. The attenuation correction function is used to limit the range of correction influence, so that the correction process is mainly concentrated in the neighborhood of the index position. Subsequently, the smoothed radius sequence is corrected according to the attenuation correction function, so that the radius value of the smoothed radius sequence at the index position gradually approaches the minimum radius value, forming a continuous transition in the neighborhood. Through the above correction process, a corrected radius sequence is formed, which maintains the overall change trend of the smoothed radius sequence and reflects the original change characteristics of the radius sequence at key positions. The corrected radius sequence is still arranged in the order of the vessel centerline data and is used in the subsequent vessel mask construction process.

[0047] The aforementioned preset conditions are used to determine whether regression correction is needed for the smoothed radius sequence at the index position. For example, in one implementation, the preset condition is satisfied when the difference is greater than zero, i.e., when the radius value of the smoothed radius sequence at the index position is greater than the minimum radius. In another implementation, the preset condition is satisfied when the difference is greater than a preset threshold, which can be set according to the overall change range of the radius sequence. In yet another implementation, the preset condition is satisfied when the ratio between the difference and the minimum radius exceeds a preset ratio. By limiting the preset conditions, the attenuation correction function is constructed only when the smoothing process causes a change in the radius at the index position, making the correction process targeted.

[0048] This application performs local regression correction on the smoothed radius sequence to maintain the original radius distribution characteristics at key positions, while maintaining the continuity of overall radius changes, which is beneficial to the consistency of subsequent vascular spatial structure expression.

[0049] In one exemplary embodiment of this application, one method for determining the centerline points to be removed and the clipping flag based on the spatial relationship between the vascular centerline data and the target area mask is as follows: mapping the physical coordinates of the centerline points in the vascular centerline data to voxel coordinates; determining whether each centerline point is located within the target area mask based on the voxel coordinates; statistically analyzing the proportion of centerline points located within the target area mask, and determining the centerline points to be removed based on the comparison result between the proportion and a preset proportion threshold; and setting the clipping flag based on the removal status of the centerline points corresponding to the starting index position in the vascular centerline data.

[0050] In this embodiment, the physical coordinates of the centerline points in the vascular centerline data are first mapped to voxel coordinates, establishing a spatial correspondence between the centerline points and the voxel space in the medical image data. Then, based on the voxel coordinates, it is determined whether each centerline point is within the target region mask. By traversing the centerline points in the vascular centerline data, the proportion of centerline points located within the target region mask is statistically analyzed. Based on the comparison between this proportion and a preset proportion threshold, centerline points to be removed are determined. This process ensures that the vascular centerline data reflects the effective spatial range of blood vessels in the medical image data. After determining the centerline points to be removed, a clipping flag is further set based on the removal status of the centerline points corresponding to the starting index position in the vascular centerline data. When the centerline point corresponding to the starting index position is determined to be a centerline point to be removed, a clipping flag is set to control the processing method of the model's end face during subsequent 3D mesh model processing.

[0051] The aforementioned preset ratio threshold is used to determine whether the distribution of centerline points within the target region mask in the current vascular centerline data requires removal. For example, in one implementation, the preset ratio threshold can be set to 80%. When the proportion of centerline points within the target region mask is less than this ratio threshold, the centerline points within the target region mask are identified as centerline points to be removed. In another implementation, the preset ratio threshold can be set based on the length of the vascular centerline or the spatial range of the target region mask. When the proportion is lower or higher than the preset ratio threshold, different centerline point processing strategies are triggered. By setting the preset ratio threshold, the determination of centerline points to be removed has a clear quantitative basis, thereby ensuring the continuity and stability of the vascular centerline data after spatial filtering.

[0052] This embodiment uses voxel space to determine the spatial relationship between the centerline point and the target area mask, thereby filtering the range of the blood vessel centerline. It also provides a basis for subsequent model end-face processing by using clipping markers, ensuring that the blood vessel modeling range is consistent with the spatial structure of the medical image.

[0053] In one exemplary embodiment of this application, one method for constructing a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed is as follows: interpolation sampling processing is performed on the vascular centerline data; the physical distance from each voxel to the centerline point in the vascular centerline data is calculated based on the voxel spacing parameter of the medical image data; the nearest centerline projection point of each voxel in the vascular centerline data is determined, and the corrected radius corresponding to the nearest centerline projection point is obtained; when the physical distance is less than the corrected radius, the voxel is marked as a vascular region voxel, and a vascular mask is constructed based on all marked vascular region voxels.

[0054] In this embodiment, interpolation sampling is first performed on the vascular centerline data to increase the sampling density of centerline points in space, making the representation of the vascular centerline in voxel space more continuous. Then, based on the voxel spacing parameter of the medical image data, the physical distance from each voxel in the voxel space to the centerline point in the vascular centerline data is calculated. During the distance calculation, the distance in different directions is uniformly measured using the voxel spacing parameter, ensuring that the distance calculation results reflect the true spatial distance relationships. For each voxel in the voxel space, its nearest centerline projection point in the vascular centerline data is determined, and the corresponding correction radius is obtained. Subsequently, based on the comparison between the physical distance between the voxel and the nearest centerline projection point and the correction radius, the voxel is used to determine the vascular region. When the physical distance is less than the correction radius, the voxel is marked as a vascular region voxel, and a vascular mask is constructed based on all marked vascular region voxels.

[0055] This embodiment determines spatial distance by combining vascular centerline data and a corrected radius sequence in voxel space, enabling the vascular region to be continuously expressed along the centerline direction, thereby forming a structurally complete vascular mask.

[0056] In one exemplary embodiment of this application, after constructing a vascular mask based on all labeled vascular region voxels, one implementation is as follows: obtain the stenosis rate value corresponding to the nearest centerline projection point; convert the stenosis rate value into an integer label value according to a preset mapping relationship, and write it into the vascular region voxel.

[0057] In this embodiment, after constructing the vascular mask, the stenosis rate value corresponding to the nearest centerline projection point is further obtained. The stenosis rate value is used to describe the structural changes at different locations along the vascular centerline. Subsequently, the stenosis rate value is converted into an integer label value according to a preset mapping relationship, and the integer label value is written into the corresponding vascular region voxel. This ensures that the voxels in the vascular mask not only contain vascular spatial distribution information but also structural attribute information that varies along the vascular centerline. Through this processing, the vascular mask forms an information set with attribute identifiers, providing support for subsequent 3D mesh model generation and model data representation.

[0058] The aforementioned pre-defined mapping relationship is used to convert continuous stenosis rate values ​​into discrete integer label values. For example, in one implementation, the stenosis rate value is multiplied by a fixed scaling factor and then converted to an integer, mapping the stenosis rate value from a decimal form to a corresponding integer form. In another implementation, the stenosis rate value is divided into intervals according to a pre-defined segmentation interval, and different intervals are assigned different integer label values. Through this mapping method, the label values ​​stored in the voxels of the vascular region can reflect the structural changes of the blood vessel along the centerline direction, while also facilitating attribute mapping processing during the 3D mesh model generation stage.

[0059] This embodiment writes the label information corresponding to the stenosis rate into the voxels of the vascular region, so that the vascular mask contains both spatial structure information and structural attribute information, which is convenient for subsequent models to express the structural changes of the blood vessel along the centerline direction.

[0060] In one exemplary embodiment of this application, an implementation method for generating a three-dimensional mesh model based on a vascular mask and performing end-face clipping processing on the three-dimensional mesh model according to clipping flags is as follows: blurring the vascular mask and generating a triangular mesh model based on an isosurface extraction algorithm; constructing a clipping plane based on the tangent vector of the centerline point at the endpoint in the vascular centerline data after removing the centerline point to be removed, and using the clipping plane to perform geometric segmentation processing on the triangular mesh model to obtain at least two mutually independent sub-mesh models; calculating the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before performing geometric segmentation processing, and retaining the sub-mesh model with the smaller distance.

[0061] In this embodiment, the vascular mask is first blurred to create a continuous transition structure in voxel space at its boundary. Then, a 3D mesh is generated based on an isosurface extraction algorithm, converting the vascular region into a triangular mesh model, allowing the vascular spatial structure to be represented in mesh form. After generating the 3D mesh model, a clipping plane is constructed based on the tangent vectors of the endpoints of the vascular centerline data after removing the centerline points to be removed. This clipping plane is then used to perform geometric segmentation on the triangular mesh model, resulting in at least two independent sub-mesh models. Subsequently, the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before geometric segmentation is calculated. The sub-mesh model with the smaller distance is retained, ensuring that the retained model corresponds to the effective structural region of the vascular vessel.

[0062] This embodiment performs geometric segmentation processing on the three-dimensional mesh model based on the clipping plane, and determines the model to be retained by combining the geometric center distance relationship, so that the range of the generated three-dimensional blood vessel structure is consistent with the corresponding area of ​​the blood vessel centerline data, thereby making the end structure expression of the model more stable.

[0063] Based on the above description of an embodiment of a three-dimensional reconstruction method for stenotic lesions based on the vascular centerline, in order to improve the problems of severe artifacts, loss of stenosis features, and tissue adhesion in existing vascular three-dimensional reconstructions, a high-fidelity three-dimensional reconstruction and visualization scheme for stenotic lesions based on the vascular centerline is introduced below. (Refer to...) Figure 2 The proposed scheme is a new framework that integrates centerline optimization, intelligent radius correction, anatomical separation and mesh generation. It consists of four parts: radius smoothing strategy, intelligent aortic separation, vascular mask construction and three-dimensional mesh generation and trimming.

[0064] First, the vessel centerline and feature data are read, and radius anchor point regression smoothing is performed to eliminate radius jitter while forcibly preserving the true values ​​at the narrowest point. Second, intelligent aortic separation is performed, automatically removing vessel points embedded in the aorta based on spatial relationships and setting clipping markers. Then, an anisotropic curved vessel mask is constructed, and anisotropic distance transformation is used to eliminate step artifacts. Finally, a 3D mesh is generated and intelligent end-face clipping is performed, outputting a high-quality VTK XML PolyData (VTP) model with quantitative color coding. Here, VTK stands for Visualization Toolkit, and XML stands for Extensible Markup Language.

[0065] The following is a detailed introduction to each part.

[0066] 1. Radius smoothing strategy based on anchor point regression This section aims to resolve the contradiction between "smoothing and denoising" and "preserving the extreme values ​​of lesions". Traditional filtering can lead to an increase in the radius of the narrowest point. This solution proposes a strategy of "smoothing first, then regression".

[0067] Step 1: Extract radius sequence and anchor points. Extract the radius sequence corresponding to the original vessel centerline from the stenosis lesion information. Point_Radii Record the minimum value and its index position in the original radius sequence, and use it as the "anchor".

[0068] Step 2: Radius Sequence Filtering. Perform median filtering and Gaussian filtering sequentially on the radius sequence. Adaptively adjust the filter window size and Gaussian kernel based on the drastic change in radius (the difference between the maximum and minimum values). SigmaThe parameters are as shown in equation (1), generating a smoothed radius sequence. Roint_Radii_Smooth .

[0069] Sigma = Point_Radii max - Point_Radii min (1) in, Point_Radii max For the maximum radius, Point_Radii min The minimum radius is expressed in millimeters.

[0070] Step 3: Perform Anchor Regression. Calculate the difference between the value of the smoothed curve at the anchor point and the original minimum value. diff .like diff>0 (i.e., smoothing causes the radius to be enlarged at narrow points), so a Gaussian decay correction function centered on the anchor point is constructed, and the correction amount is subtracted from the smooth curve. The corrected formula is approximately Equation (2).

[0071] (2) in, sigma Control the scope of the correction's impact. x For the first x One central point, min_idx The index of the point with the smallest radius.

[0072] 2. Intelligent separation and marker setting of the aortic region This section is used to automatically handle adhesions between the origin of coronary arteries, carotid arteries, and other branch vessels and the aorta.

[0073] Step 1: Determine the center point location. Input the aorta segmentation mask (Aorta Label) of the original image, traverse the center line points of the blood vessels, map the physical coordinates of the points back to voxel coordinates, and determine whether they are located within the aorta mask.

[0074] Step 2: Determine the location of the stenosis lesion. Calculate the proportion of points located inside the aorta. If the proportion exceeds the threshold (set to 80%), the vessel segment is determined to be inside the aorta, and the illegal removal operation is stopped; otherwise, the stenosis lesion is determined to be located on a branch vessel, and all points located inside the aorta are marked.

[0075] Step 3: Morphological expansion removal. This involves not only removing points located within the aorta, but also further removing adjacent points anteriorly / posteriorly. N points ( N Set to 4) to ensure that the origin of the blood vessel is completely detached from the aortic wall.

[0076] Step 4: Dynamically update the clipping flag. To facilitate the separation of branch vessels from the aorta, the starting end of the branch vessel is set to a closed tubular shape, while the remaining vessels are set to hollow tubular shapes. If the physical starting point (index 0) of the branch vessel is removed, it indicates that the current starting point is the true starting point of the branch vessel, and the clipping flag Start0_Flag is set to 1. Simultaneously, the stenosis rate list at each time point is updated synchronously. Sten_Rate_List This ensures consistency between the narrowing rate list and the centerline length.

[0077] 3. Construction of anisotropic tortuous vascular mask This section addresses the issue of step-like artifacts by constructing a high-precision vascular geometry mask in voxel space.

[0078] Step 1: Draw the vessel centerline. Draw the vessel centerline in voxel space, and increase the sampling density using linear interpolation (set to 2 times the sampling) to ensure that the centerline is continuous and without breaks.

[0079] Step 2: Perform anisotropic Euclidean distance transformation. When calculating the distance from a voxel to the center line, spatial parameters of the image are introduced to calculate the true physical distance.

[0080] Step 3: Generate a label mask. For each voxel in the image, find its nearest projection point on the center line and obtain the smooth radius of that point. If the physical distance from the voxel to the center line is less than this radius, it is marked as a blood vessel region. Obtain the stenosis rate value corresponding to that projection point (…). Stenosis Rate ), using equation (3) to convert the integer label values Voxel_Value The linear mapping is converted into integer label values ​​and written into the voxel. Through this step, the generated voxel mask is no longer a single binary image, but an "attribute field" carrying quantitative pathological information distributed along the vessel axis, providing direct data support for the subsequent generation of color-coded 3D models.

[0081] Voxel_Value=int(Stenosis_Rate*100) (3) Here, int() represents integer conversion operation, and 100 is the scaling factor.

[0082] Step 4: End face flattening. Construct a normal plane using the tangent vectors at the endpoints of the centerline, and use the dot product method to remove the hemispherical protrusions at both ends of the blood vessel, making the mask ends flat and perpendicular to the centerline.

[0083] 4. 3D Mesh Generation and Trimming (VTP Generation) This section converts the voxel mask into a vector model (VTP format) that is easy to render and interact with.

[0084] Step 1: Image and mesh smoothing. Gaussian blur preprocessing is applied to the binary mask ( Sigma=0.8 Then, the Marching Cubes algorithm (a 3D mesh generation algorithm for extracting isosurfaces from 3D voxel data) is used to extract isosurfaces and generate triangular meshes, which can significantly reduce the step effect on the mesh surface.

[0085] Step 2: Attribute Mapping. Using dilatation sampling, the values ​​in the original mask (representing the narrowing rate or sequence ID) are mapped to the mesh vertices, and a red-white gradient color map is generated based on the values ​​to visually display the degree of narrowing.

[0086] Step 3: Intelligent Clipping Based on Flags. Determine if model endface trimming is needed based on Start0_Flag. If necessary, construct a VTK clipping plane (vtkPlane) based on the endpoint tangent vector. Key error prevention logic: Employ a "preserve the proximal end" strategy. After clipping, calculate the distance from the geometric center of the preserved portion and the excised portion to the overall vessel center, always preserving the portion with the closer distance to prevent "reverse cutting" due to incorrect normal direction calculation.

[0087] Step 4: Encapsulate the output. Encapsulate the vertex coordinates, patch indices, scalar attribute identifier values ​​(LabelI) associated with the vertices or cells of the 3D mesh model, and RGB color vectors into a VTP format file, and finally output the narrow lesion VTP model.

[0088] Based on the above embodiments, the following describes a scenario for three-dimensional modeling of branch vascular structures using medical imaging data from computed tomography angiography (CTA). First, the CTA medical imaging data of the object to be processed is imported into the vascular modeling processing system. Vascular region extraction processing is then performed on the CTA medical imaging data to obtain vascular centerline data and the radius sequence corresponding to each centerline point.

[0089] Subsequently, data preprocessing is performed on the radius sequence. Specifically, the radius sequence is traversed and analyzed to determine the minimum radius and its corresponding index position. Then, filtering is applied to the radius sequence to obtain a smoothed radius sequence. During filtering, filtering parameters are determined based on the difference between the maximum and minimum radii in the radius sequence. Median filtering is used to adjust for local abnormal fluctuations in the radius sequence, and Gaussian filtering is used for continuous smoothing. After filtering, regression correction is performed based on the minimum radius and the smoothed radius sequence. By calculating the radius difference at the index position and constructing a decay correction function, local correction is applied to the smoothed radius sequence to obtain the corrected radius sequence.

[0090] After radius data correction, spatial filtering is performed on the vessel centerline data. First, a target region mask is obtained from the CTA medical image data, and the physical coordinates of the centerline points in the vessel centerline data are mapped to voxel coordinates from the CTA image data. Then, by determining whether each centerline point is within the target region mask, the proportion of centerline points within the target region mask is statistically analyzed, and the centerline points to be removed are determined based on a comparison with a preset proportion threshold. After determining the centerline points to be removed, a clipping flag is set based on the removal status of the centerline points corresponding to the starting index position in the vessel centerline data to record the spatial validity of the vessel centerline's starting end.

[0091] After obtaining the vessel centerline data and the corrected radius sequence after removing the centerline points to be removed, a vessel mask is constructed in the voxel space corresponding to the CTA image data. Specifically, interpolation sampling is performed on the vessel centerline data to increase the distribution density of centerline points in space. Subsequently, based on the voxel spacing parameters in the CTA image data, the physical distance from each voxel in the voxel space to the centerline point in the vessel centerline data is calculated. For each voxel, its nearest centerline projection point in the vessel centerline data is determined, and the corrected radius corresponding to that projection point is obtained. When the physical distance from a voxel to the nearest centerline projection point is less than the corrected radius, the voxel is marked as a vessel region voxel, and a vessel mask is constructed based on all vessel region voxels. Simultaneously, after constructing the vessel mask, the stenosis rate value corresponding to the nearest centerline projection point is converted into an integer label value and written into the vessel region voxel, so that the vessel mask contains structural attribute information.

[0092] After constructing the vascular mask, a 3D model generation process is performed. First, the vascular mask is blurred to reduce voxel boundary discretization. Then, a triangular mesh model is generated based on an isosurface extraction algorithm, allowing the vascular structure to be represented in a polygonal mesh format. Next, a clipping plane is constructed based on clipping flags and the tangent vectors of the centerline points at the endpoints in the vascular centerline data. This clipping plane is then used to perform geometric segmentation on the triangular mesh model, resulting in at least two independent sub-mesh models. Next, the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before geometric segmentation is calculated, and the sub-mesh model with the smaller distance is retained. Finally, the vertex coordinates, face indices, LabelIDs, and RGB color data of the retained triangular mesh models are encapsulated into a VTP model file for vascular structure display and subsequent data processing.

[0093] Through the above processing, vascular spatial structure modeling can be completed under CTA medical imaging data, and data model results that can be used to display vascular structures can be generated.

[0094] It should be noted that the three-dimensional reconstruction method for stenotic lesions based on the vascular centerline provided in this application can be executed by a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline, or by a control module within that system for executing the three-dimensional reconstruction method. This application uses the execution of the three-dimensional reconstruction method for stenotic lesions based on the vascular centerline by a three-dimensional reconstruction system for stenotic lesions as an example to illustrate the three-dimensional reconstruction method for stenotic lesions based on the vascular centerline provided in this application.

[0095] Reference Figure 3 This illustration shows a schematic diagram of a three-dimensional reconstruction system for stenotic lesions based on the vascular centerline according to an embodiment of this application. The system may specifically include the following modules: Data acquisition module 31 is used to acquire vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data; The smooth radius sequence determination module 32 is used to determine the minimum radius value and the corresponding index position based on the radius sequence, and to filter the radius sequence to obtain a smooth radius sequence; The modified radius sequence determination module 33 is used to perform regression correction processing based on the minimum radius value and the smoothed radius sequence to obtain the modified radius sequence; The line point marker determination module 34 is used to obtain a target area mask based on the medical image data, and determine the center line points to be removed and the clipping markers based on the spatial positional relationship between the blood vessel center line data and the target area mask. The vascular mask construction module 35 is used to construct a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed. The trimming module 36 is used to generate a three-dimensional mesh model based on the blood vessel mask, and to perform end-face trimming processing on the three-dimensional mesh model according to the trimming flag.

[0096] In one exemplary embodiment of this application, the smooth radius sequence determination module 32 includes: A radius acquisition module is used to acquire the maximum and minimum radii of the radius sequence. A filter parameter determination module is used to determine filter parameters based on the difference between the maximum radius and the minimum radius; The filtering module is used to sequentially perform median filtering and Gaussian filtering on the radius sequence to obtain the smoothed radius sequence.

[0097] In one exemplary embodiment of this application, the modified radius sequence determination module 33 includes: The difference calculation module is used to calculate the difference between the radius value of the smoothed radius sequence at the index position and the minimum radius value; The function construction module is used to construct a decay correction function centered on the index position when the difference meets a preset condition; The smooth radius correction module is used to correct the smooth radius sequence according to the decay correction function to obtain the corrected radius sequence.

[0098] In one exemplary embodiment of this application, the line dot flag determination module 34 includes: The coordinate mapping module is used to map the physical coordinates of the centerline points in the blood vessel centerline data to voxel coordinates. The coordinate determination module is used to determine whether each centerline point is located within the target area mask based on the voxel coordinates. The centerline point determination module is used to statistically analyze the proportion of centerline points located within the target area mask, and determine the centerline points to be removed based on the comparison result between the proportion and a preset proportion threshold. The flag setting module is used to set the clipping flag based on the removal status of the centerline point corresponding to the starting index position in the vascular centerline data.

[0099] In one exemplary embodiment of this application, the vascular mask construction module 35 includes: An interpolation sampling module is used to perform interpolation sampling processing on the blood vessel centerline data; The physical distance calculation module is used to calculate the physical distance from each voxel to the centerline point in the blood vessel centerline data based on the voxel spacing parameter of the medical image data. The correction radius determination module is used to determine the nearest centerline projection point of each voxel in the blood vessel centerline data, and obtain the correction radius corresponding to the nearest centerline projection point; A vascular region voxel marking module is used to mark the voxel as a vascular region voxel when the physical distance is less than the correction radius, and to construct the vascular mask based on all the marked vascular region voxels.

[0100] In one exemplary embodiment of this application, the system further includes: The stenosis rate acquisition module is used to acquire the stenosis rate value corresponding to the nearest centerline projection point after the vascular region voxel marking module constructs the vascular mask based on all marked vascular region voxels. The numerical conversion module is used to convert the stenosis rate value into an integer label value according to a preset mapping relationship and write it into the vascular region voxel.

[0101] In one exemplary embodiment of this application, the cropping processing module 36 includes: A triangular mesh model generation module is used to blur the blood vessel mask and generate a triangular mesh model based on an isosurface extraction algorithm. The sub-mesh model determination module is used to construct a clipping plane based on the tangent vector of the centerline point at the endpoint in the blood vessel centerline data after removing the centerline point to be removed, and to use the clipping plane to perform geometric segmentation processing on the triangular mesh model to obtain at least two independent sub-mesh models. The sub-mesh model retention module is used to calculate the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before the geometric segmentation process is performed, and to retain the sub-mesh model with the smaller distance.

[0102] The three-dimensional reconstruction system for stenotic lesions based on the vascular centerline in this application embodiment can be a device, or a component, integrated circuit, or chip in a terminal. The device can be a mobile electronic device or a non-mobile electronic device. For example, mobile electronic devices can be mobile phones, tablets, laptops, PDAs, in-vehicle electronic devices, wearable devices, ultra-mobile personal computers (UMPCs), netbooks, or personal digital assistants (PDAs), etc., while non-mobile electronic devices can be servers, network-attached storage (NAS), personal computers (PCs), televisions (TeleVision), ATMs, or self-service machines, etc. This application embodiment does not impose specific limitations.

[0103] The three-dimensional reconstruction system for stenotic lesions based on the vascular centerline in this application embodiment can be a device with an operating system. This operating system can be Android, iOS, or other possible operating systems; this application embodiment does not specifically limit it.

[0104] The three-dimensional reconstruction system for stenotic lesions based on the vascular centerline provided in this application embodiment can achieve… Figure 1 The various processes implemented in the method embodiments of the three-dimensional reconstruction system for stenotic lesions based on the vascular centerline are not described in detail here to avoid repetition.

[0105] Optionally, embodiments of this application also provide an electronic device, including a processor, a memory, and a program or instructions stored in the memory and executable on the processor. When the program or instructions are executed by the processor, they implement the various processes of the above-described three-dimensional reconstruction method for stenotic lesions based on the vascular centerline and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0106] It should be noted that the electronic devices in the embodiments of this application include the mobile electronic devices and non-mobile electronic devices described above.

[0107] Figure 4 A schematic diagram of the hardware structure of an electronic device to implement an embodiment of this application.

[0108] The electronic device 1000 includes, but is not limited to, components such as: a radio frequency unit 1001, a network module 1002, an audio output unit 1003, an input unit 1004, a sensor 1005, a display unit 1006, a user input unit 1007, an interface unit 1008, a memory 1009, and a processor 1010. The input unit 1004 may include a graphics processor 10041 and a microphone 10042. The display unit 1006 may include a display panel 10061. The user input unit 1007 may include a touch panel 10071 and other input devices 10072. The memory 1009 may include applications and an operating system.

[0109] Those skilled in the art will understand that the electronic device 1000 may also include a power supply (such as a battery) for supplying power to various components. The power supply may be logically connected to the processor 1010 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system. Figure 4 The electronic device structure shown does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements, which will not be elaborated here.

[0110] This application also provides a readable storage medium storing a program or instructions. When the program or instructions are executed by a processor, they implement the various processes of the above-described three-dimensional reconstruction method for stenotic lesions based on the vascular centerline and achieve the same technical effect. To avoid repetition, they will not be described again here.

[0111] The processor mentioned above is the processor in the electronic device described in the above embodiments. The readable storage medium includes computer-readable storage media, such as computer read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk.

[0112] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and systems in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0113] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal (which may be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0114] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A method for three-dimensional reconstruction of stenotic lesions based on the vascular centerline, characterized in that, The method includes: Acquire vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data; The minimum radius and its corresponding index position are determined based on the radius sequence, and the radius sequence is filtered to obtain a smooth radius sequence. Based on the minimum radius value and the smoothed radius sequence, a regression correction process is performed to obtain the corrected radius sequence; The target region mask is obtained based on the medical image data, and the centerline point to be removed and the clipping marker are determined based on the spatial relationship between the blood vessel centerline data and the target region mask. Based on the corrected radius sequence and the vessel centerline data after removing the centerline points to be removed, a vessel mask is constructed in voxel space. A three-dimensional mesh model is generated based on the blood vessel mask, and the end face of the three-dimensional mesh model is clipped according to the clipping flag.

2. The method according to claim 1, characterized in that, The step of filtering the radius sequence to obtain a smooth radius sequence includes: Obtain the maximum and minimum radii of the radius sequence; The filtering parameters are determined based on the difference between the maximum radius and the minimum radius; The radius sequence is processed sequentially by median filtering and Gaussian filtering to obtain the smoothed radius sequence.

3. The method according to claim 1, characterized in that, The step of performing regression correction processing based on the minimum radius value and the smoothed radius sequence to obtain the corrected radius sequence includes: Calculate the difference between the radius value of the smoothed radius sequence at the index position and the minimum radius value; When the difference meets the preset conditions, a decay correction function centered on the index position is constructed; The smooth radius sequence is corrected according to the attenuation correction function to obtain the corrected radius sequence.

4. The method according to claim 1, characterized in that, The step of determining the centerline point to be removed and the clipping marker based on the spatial relationship between the blood vessel centerline data and the target area mask includes: Map the physical coordinates of the centerline points in the blood vessel centerline data to voxel coordinates; Determine whether each centerline point is located within the target region mask based on the voxel coordinates; The proportion of centerline points located within the target area mask is statistically analyzed, and the centerline points to be removed are determined based on the comparison between the proportion and a preset proportion threshold. The clipping flag is set based on the removal status of the centerline point corresponding to the starting index position in the vascular centerline data.

5. The method according to claim 1, characterized in that, The step of constructing a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed includes: The blood vessel centerline data are subjected to interpolation sampling processing; Based on the voxel spacing parameters of the medical image data, calculate the physical distance of each voxel to the centerline point in the vascular centerline data; Determine the nearest centerline projection point for each voxel in the vascular centerline data, and obtain the correction radius corresponding to the nearest centerline projection point; When the physical distance is less than the corrected radius, the voxel is marked as a vascular region voxel, and the vascular mask is constructed based on all the marked vascular region voxels.

6. The method according to claim 5, characterized in that, After constructing the vascular mask based on all marked vascular region voxels, the method further includes: Obtain the narrowing rate value corresponding to the nearest centerline projection point; The stenosis rate value is converted into an integer label value according to a preset mapping relationship and written into the vascular region voxel.

7. The method according to claim 1, characterized in that, The step of generating a three-dimensional mesh model based on the vascular mask and performing end-face clipping processing on the three-dimensional mesh model according to the clipping flags includes: The blood vessel mask is blurred, and a triangular mesh model is generated based on the isosurface extraction algorithm; Based on the tangent vector of the centerline point at the endpoint in the blood vessel centerline data after removing the centerline point to be removed, a clipping plane is constructed, and the clipping plane is used to perform geometric segmentation on the triangular mesh model to obtain at least two independent sub-mesh models. Calculate the distance between the geometric center of each sub-mesh model and the geometric center of the triangular mesh model before performing the geometric segmentation process, and retain the sub-mesh model with the smaller distance.

8. A three-dimensional reconstruction system for stenotic lesions based on the vascular centerline, characterized in that, The system includes: The data acquisition module is used to acquire vascular centerline data, the radius sequence corresponding to the centerline, and medical imaging data; A smooth radius sequence determination module is used to determine the minimum radius value and its corresponding index position based on the radius sequence, and to filter the radius sequence to obtain a smooth radius sequence; The modified radius sequence determination module is used to perform regression correction processing based on the minimum radius value and the smoothed radius sequence to obtain the modified radius sequence; The line point marker determination module is used to obtain a target area mask based on the medical image data, and determine the center line points to be removed and the clipping markers based on the spatial positional relationship between the blood vessel centerline data and the target area mask. A vascular mask construction module is used to construct a vascular mask in voxel space based on the corrected radius sequence and the vascular centerline data after removing the centerline points to be removed. The trimming module is used to generate a three-dimensional mesh model based on the blood vessel mask, and to perform end-face trimming processing on the three-dimensional mesh model according to the trimming flag.

9. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the three-dimensional reconstruction method for stenotic lesions based on the vascular centerline as described in any one of claims 1-7.

10. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the three-dimensional reconstruction method for stenotic lesions based on the vascular centerline as described in any one of claims 1-7.

Images