A method for evaluating vascular stenosis based on vascular geometric attribute analysis

Abstract

The application discloses a kind of based on blood vessel geometric attribute analysis blood vessel stenosis evaluation method, it is related to medical image processing technical field, including: obtaining target blood vessel image and extracting blood vessel cavity model;Extract blood vessel center line, establish multiple sections along center line, obtain section geometric data and wall radial data;According to the geometric continuity of adjacent section, generate section validity sequence, and the original actual cavity diameter, section area and wall radial data are corrected with credibility;Based on proximal, distal reference blood vessel segment constructs reference cavity diameter curve, reference equivalent section area sequence and reference wall radial sequence;Construct multi-dimensional cavity deviation, determine true cavity contraction position and contraction form according to its consistency, and determine stenosis action interval, output stenosis evaluation result;Reduce the influence of curved oblique cut, profile distortion and reference benchmark deviation on evaluation result, improve the stability and accuracy of blood vessel stenosis range identification and stenosis degree evaluation.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, specifically to a method for assessing vascular stenosis based on the analysis of vascular geometric properties. Background Technology

[0002] Vascular stenosis is an important imaging manifestation of cardiovascular and cerebrovascular diseases, peripheral vascular diseases, and various hemodynamic abnormalities. With the development of medical imaging technology, imaging data such as CTA, MRA, DSA, and three-dimensional vascular reconstruction can more intuitively reflect the morphology of vascular lumens, providing a data foundation for the identification, measurement, and assessment of vascular stenosis. Current methods for assessing vascular stenosis typically extract the vascular centerline based on vascular segmentation results, establish a cross-section along the centerline, and determine the degree of vascular stenosis by calculating geometric parameters such as cross-sectional area, equivalent lumen diameter, minimum lumen diameter, or lumen diameter reduction ratio.

[0003] However, in actual vascular imaging, blood vessels are not ideal straight structures. Target vessels often exhibit tortuosity, bifurcation, natural taper, local dilation, eccentric contraction, and blurred image boundaries. Existing methods, when establishing cross-sections along the vessel's centerline, may result in oblique cuts or contour distortions if the local centerline direction changes significantly, leading to inaccuracies in geometric data such as cross-sectional area, equivalent lumen diameter, and wall distance. Directly using these distorted data for stenosis assessment can easily misinterpret local cross-sectional anomalies, segmentation burrs, or projection changes caused by tortuosity as true stenosis.

[0004] Meanwhile, existing methods typically use a single reference section at the proximal or distal end of the lesion, or simple interpolation between proximal and distal reference values, when constructing a normal reference lumen diameter. This approach fails to adequately reflect the continuous changes in the target vessel along the centerline. When the vessel being evaluated is located in a naturally tapering segment, a tortuous segment, or a region with poor reference section stability, the reference lumen diameter is prone to being too large or too small, thus affecting the reliability of the stenosis calculation results.

[0005] Furthermore, vascular stenosis not only manifests as a reduction in lumen diameter or cross-sectional area, but may also present as eccentric wall constriction. For eccentric stenosis, relying solely on equivalent lumen diameter or cross-sectional area makes it difficult to distinguish between true wall constriction and pseudo-constriction caused by local contour disturbances, segmentation errors, or abnormal cross-sectional orientation. Existing methods often lack a comprehensive assessment of the consistency between changes in lumen diameter, cross-sectional area, and radial changes in the wall, resulting in insufficient stability in the evaluation results of stenosis location, stenosis boundaries, and stenosis morphology. Summary of the Invention

[0006] In view of the shortcomings of the prior art described above, the purpose of this invention is to provide a method for assessing vascular stenosis based on the analysis of vascular geometric properties, so as to solve the above-mentioned technical problems.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a method for assessing vascular stenosis based on vascular geometric property analysis, comprising:

[0008] Acquire images of the target blood vessel and extract a model of the target blood vessel cavity;

[0009] Extract the centerline of the target blood vessel, set multiple sampling positions along the centerline of the target blood vessel, establish corresponding cross sections at each sampling position, and obtain the cross section geometric data and wall radial data of each cross section;

[0010] Based on the geometric continuity between each cross section and adjacent cross sections, a cross section validity sequence distributed along the centerline of the target blood vessel is generated. Based on the cross section validity sequence, the credibility of the original actual lumen diameter, original actual cross-sectional area and wall radial data corresponding to each sampling position is corrected to obtain the corrected actual lumen diameter sequence, corrected actual cross-sectional area sequence and corrected wall radial data sequence.

[0011] The vascular interval to be analyzed is determined on the centerline of the target vascular segment. Based on the cross-sectional validity sequence and the corrected actual lumen diameter sequence, the proximal reference vascular segment and the distal reference vascular segment are determined at the proximal and distal ends of the vascular interval to be analyzed, respectively. The reference lumen diameter curve is constructed based on the corresponding lumen diameter data of the proximal and distal reference vascular segments. The reference equivalent cross-sectional area sequence is generated based on the reference lumen diameter curve. The reference wall radial sequence is generated based on the reference lumen diameter curve and the corrected wall radial data of the proximal and distal reference vascular segments.

[0012] Based on correcting the differences between the actual cavity diameter sequence and the reference cavity diameter curve, correcting the differences between the actual cross-sectional area sequence and the reference equivalent cross-sectional area sequence, and correcting the differences between the wall radial data sequence and the reference wall radial data sequence, a multidimensional cavity deviation is constructed, including the first cavity deviation, the second cavity deviation, and the third cavity deviation.

[0013] Based on the consistency relationship between the multidimensional cavity deviations, the corrected actual cavity diameter, corrected actual cross-sectional area, and corrected wall radial data of the sampling positions that do not meet the preset consistency relationship are updated. The actual cavity contraction position and contraction shape are determined based on the multidimensional cavity deviations that meet the preset consistency relationship.

[0014] The stenosis range is determined based on the contraction locations of multiple consecutive real cavities, and the stenosis assessment result is output based on the stenosis range.

[0015] The present invention is further configured to establish corresponding cross-sections at each sampling location, and acquire cross-sectional geometric data and wall radial data of each cross-section, including:

[0016] Based on the local tangential direction of the target blood vessel centerline at the current sampling position, establish a cross-sectional plane orthogonal to the local tangential direction;

[0017] Intersecting the cross-sectional plane with the target blood vessel cavity model yields the cross-sectional profile corresponding to the current sampling position;

[0018] Based on the cross-sectional profile, the cross-sectional area, equivalent cavity diameter, cross-sectional center, cross-sectional perimeter, cross-sectional major axis length, and cross-sectional minor axis length are obtained as cross-sectional geometric data;

[0019] Using the center of the cross section as the radial starting point, the distance from the center of the cross section to the cross section profile is obtained along multiple preset radial directions to obtain the radial data of the wall surface.

[0020] The present invention is further configured to generate a sequence of effective sections distributed along the centerline of the target blood vessel based on the geometric continuity relationship between each section and adjacent sections, including:

[0021] For the cross section corresponding to the i-th sampling position, calculate the area continuity deviation between the cross section area and the cross section area of ​​the adjacent cross section;

[0022] Calculate the center offset deviation between the center of this section and the center of the adjacent section;

[0023] Calculate the directional deviation between the centerline direction at the sampling location and the centerline direction at the adjacent sampling location;

[0024] Based on area continuity deviation, center offset deviation, and direction deflection deviation, the effective value of the cross section corresponding to the i-th sampling position is generated;

[0025] The effective values ​​of the cross sections corresponding to multiple sampling locations are arranged in the sampling order of the target blood vessel centerline to obtain the cross section effectiveness sequence.

[0026] The present invention is further configured to generate the effective cross-sectional value corresponding to the i-th sampling position based on area continuity deviation, center offset deviation, and direction deflection deviation, including:

[0027] The area continuity deviation, center offset deviation, and direction deflection deviation are compared with their respective preset deviation ranges.

[0028] When the area continuity deviation, center offset deviation, and direction deflection deviation are all within the corresponding first deviation range, the cross section corresponding to the i-th sampling position is determined as the high effective cross section, and the first effective value of the cross section is generated.

[0029] When any one of the area continuous deviation, center offset deviation, and direction deflection deviation is within the corresponding second deviation range, and there is no deviation within the corresponding third deviation range, the cross section corresponding to the i-th sampling position is determined as the effective cross section, and the effective value of the second cross section is generated.

[0030] When any one of the area continuity deviation, center offset deviation, and direction deflection deviation falls within the corresponding third deviation range, the cross section corresponding to the i-th sampling position is determined as the low effective cross section, and the effective value of the third cross section is generated; wherein, the first deviation range, the second deviation range, and the third deviation range are set in order of increasing deviation degree, and the effective value of the first cross section, the effective value of the second cross section, and the effective value of the third cross section are set in order of decreasing cross section confidence degree;

[0031] The effective values ​​of the cross sections corresponding to multiple sampling locations are arranged in the sampling order of the target blood vessel centerline to obtain the cross section effectiveness sequence.

[0032] The present invention is further configured to perform credibility correction on the original actual cavity diameter, original actual cross-sectional area, and wall radial data corresponding to each sampling position based on the cross-sectional validity sequence, including:

[0033] The cross-sectional acceptance weight of each sampling location is determined based on the effective value of the cross-section at each sampling location.

[0034] Based on the cross-sectional effective values ​​of adjacent sampling positions that meet the preset effective conditions, generate the neighboring extended cavity diameter, neighboring extended cross-sectional area, and neighboring extended wall radial data corresponding to the current sampling position;

[0035] Based on the cross-sectional acceptance weight, the original actual cavity diameter at the current sampling position is fused with the neighboring extended cavity diameter to obtain the corrected actual cavity diameter;

[0036] The original actual cross-sectional area at the current sampling location is fused with the neighboring extended cross-sectional area to obtain the corrected actual cross-sectional area;

[0037] The wall radial data at the current sampling location is fused with the neighboring extended wall radial data to obtain the corrected wall radial data.

[0038] The present invention is further configured to determine the vascular region to be analyzed on the center line of the target vascular vessel, including:

[0039] Based on the corrected actual lumen diameter sequence, segments along the centerline of the target blood vessel that are continuously lower than the lumen diameter variation trend of the neighboring region are identified;

[0040] The segments whose diameter changes continuously below the trend of neighboring vascular segments are identified as the vascular intervals to be analyzed.

[0041] The present invention is further configured to determine a proximal reference vascular segment and a distal reference vascular segment at the proximal and distal ends of the vascular region to be analyzed, respectively, including:

[0042] Search for continuous vascular segments in both the proximal and distal directions of the vascular region to be analyzed;

[0043] Continuous vascular segments whose effective cross-sectional values ​​meet preset effective conditions, whose corrected actual lumen diameter changes meet preset stable conditions, and whose corrected actual cross-sectional area changes meet preset stable conditions are respectively identified as proximal reference vascular segments and distal reference vascular segments; wherein, both proximal and distal reference vascular segments are located outside the vascular interval to be analyzed.

[0044] The present invention is further configured to construct a reference cavity diameter curve, a reference equivalent cross-sectional area sequence, and a reference wall radial sequence, including:

[0045] Based on the lumen diameter variation trends of the proximal and distal reference vessel segments in the corrected actual lumen diameter sequence, a reference lumen diameter curve is constructed along the vessel interval to be analyzed.

[0046] Based on the reference cavity diameter at the corresponding position of the reference cavity diameter curve, generate a reference equivalent cross-sectional area sequence;

[0047] Based on the changing trends of the corrected wall radial data in multiple radial directions of the proximal and distal reference vessel segments, and combined with the reference lumen diameter at the corresponding position of the reference lumen diameter curve, reference wall radial data at each sampling position within the vessel segment to be analyzed are generated, forming a reference wall radial sequence.

[0048] The present invention is further configured such that the first cavity deviation is used to characterize the degree of cavity diameter reduction relative to the reference cavity diameter curve;

[0049] The second cavity deviation is used to characterize the degree of area reduction of the corrected actual cross-sectional area relative to the reference equivalent cross-sectional area sequence;

[0050] The third cavity deviation is used to characterize the degree of wall indentation of the corrected wall radial data relative to the reference wall radial sequence;

[0051] When the deviation of the first cavity and the deviation of the second cavity at the same sampling position both meet the preset contraction condition, and the difference between the deviation of the first cavity and the deviation of the second cavity is within the preset consistency range, it is determined that the sampling position meets the preset consistency relationship.

[0052] The present invention is further configured to determine the actual cavity contraction position and contraction morphology based on the multidimensional cavity deviation that satisfies a preset consistency relationship, including:

[0053] When the sampling location meets the preset consistency relationship, the sampling location is determined as the actual cavity contraction location;

[0054] When the deviation of the third cavity corresponding to the actual cavity contraction position is manifested as unilateral wall retraction in the relative radial direction, the contraction pattern of the actual cavity contraction position is determined to be an eccentric contraction pattern.

[0055] When the deviation of the third cavity corresponding to the actual cavity contraction position is manifested as wall inward contraction in multiple radial directions, the contraction pattern of the actual cavity contraction position is determined to be the centripetal contraction pattern.

[0056] The contraction locations of multiple real cavities continuously distributed along the centerline of the target blood vessel are determined as the stenosis range. Based on the minimum corrected actual lumen diameter, minimum corrected actual cross-sectional area, stenosis range length, and lumen deviation distribution within the stenosis range, the vascular stenosis assessment result is output.

[0057] This invention provides a method for assessing vascular stenosis based on vascular geometric property analysis. It involves extracting a target vascular cavity model and the target vascular centerline, establishing multiple cross-sections along the centerline, and acquiring cross-sectional geometric data and wall radial data. A cross-sectional validity sequence is generated based on the geometric continuity between adjacent cross-sections, and the original actual lumen diameter, original actual cross-sectional area, and wall radial data are then corrected for reliability, resulting in a corrected actual lumen diameter sequence, a corrected actual cross-sectional area sequence, and a corrected wall radial data sequence. A reference lumen diameter curve, a reference equivalent cross-sectional area sequence, and a reference wall radial sequence are constructed based on reference vascular segments at the proximal and distal ends of the vascular region to be analyzed. Furthermore, a multidimensional lumen deviation is constructed based on the difference between the corrected data and the reference data, and the true lumen contraction position and contraction morphology are determined based on the consistency relationship between the multidimensional lumen deviations. This determines the stenosis range and outputs the vascular stenosis assessment result. The beneficial effects include:

[0058] 1. By constructing a cross-sectional validity sequence and performing credibility correction on the original actual lumen diameter, original actual cross-sectional area, and wall radial data, the influence of vessel tortuosity, cross-sectional oblique cutting, contour distortion, or local segmentation abnormalities on cross-sectional geometric data can be reduced, thereby improving the reliability of the basic data used for subsequent stenosis assessment.

[0059] 2. By constructing reference lumen diameter curves, reference equivalent cross-sectional area sequences, and reference wall radial sequences based on proximal and distal reference vessel segments, the reference benchmark deviation caused by using only a single reference section can be avoided, making the reference data of the vessel section to be analyzed more consistent with the continuous change characteristics of the target vessel and improving the stability of stenosis determination.

[0060] 3. By constructing the first cavity deviation, the second cavity deviation, and the third cavity deviation, and determining the true cavity contraction position and contraction pattern based on the consistency relationship between the multidimensional cavity deviations, it is possible to comprehensively distinguish between true stenosis, eccentric contraction, and local pseudo-contraction, reduce misjudgments caused by single cavity diameter or cross-sectional area judgments, and improve the accuracy of stenosis action zone identification and stenosis assessment results.

[0061] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0062] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0063] Figure 1 A flowchart illustrating a method for assessing vascular stenosis based on vascular geometric property analysis, as shown in an exemplary embodiment of the present invention;

[0064] Figure 2 A schematic diagram of a target blood vessel image and a blood vessel cavity model is shown as an exemplary embodiment of the present invention;

[0065] Figure 3 A schematic diagram of the target blood vessel centerline and continuous blood vessel structure shown in an exemplary embodiment of the present invention;

[0066] Figure 4 A schematic diagram illustrating the determination of the vascular region to be analyzed, as shown in an exemplary embodiment of the present invention;

[0067] Figure 5 This is a schematic diagram of a proximal reference vessel segment and a distal reference vessel segment, illustrating an exemplary embodiment of the present invention. Detailed Implementation

[0068] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0069] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0070] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0071] A method for assessing vascular stenosis based on the analysis of vascular geometric properties, such as Figure 1 As shown, it includes:

[0072] Acquire images of the target blood vessel and extract a model of the target blood vessel cavity;

[0073] Extract the centerline of the target blood vessel, set multiple sampling positions along the centerline of the target blood vessel, establish corresponding cross sections at each sampling position, and obtain the cross section geometric data and wall radial data of each cross section;

[0074] Based on the geometric continuity between each cross section and adjacent cross sections, a cross section validity sequence distributed along the centerline of the target blood vessel is generated. Based on the cross section validity sequence, the credibility of the original actual lumen diameter, original actual cross-sectional area and wall radial data corresponding to each sampling position is corrected to obtain the corrected actual lumen diameter sequence, corrected actual cross-sectional area sequence and corrected wall radial data sequence.

[0075] The vascular interval to be analyzed is determined on the centerline of the target vascular segment. Based on the cross-sectional validity sequence and the corrected actual lumen diameter sequence, the proximal reference vascular segment and the distal reference vascular segment are determined at the proximal and distal ends of the vascular interval to be analyzed, respectively. The reference lumen diameter curve is constructed based on the corresponding lumen diameter data of the proximal and distal reference vascular segments. The reference equivalent cross-sectional area sequence is generated based on the reference lumen diameter curve. The reference wall radial sequence is generated based on the reference lumen diameter curve and the corrected wall radial data of the proximal and distal reference vascular segments.

[0076] Based on correcting the differences between the actual cavity diameter sequence and the reference cavity diameter curve, correcting the differences between the actual cross-sectional area sequence and the reference equivalent cross-sectional area sequence, and correcting the differences between the wall radial data sequence and the reference wall radial data sequence, a multidimensional cavity deviation is constructed, including the first cavity deviation, the second cavity deviation, and the third cavity deviation.

[0077] Based on the consistency relationship between the multidimensional cavity deviations, the corrected actual cavity diameter, corrected actual cross-sectional area, and corrected wall radial data of the sampling positions that do not meet the preset consistency relationship are updated. The actual cavity contraction position and contraction shape are determined based on the multidimensional cavity deviations that meet the preset consistency relationship.

[0078] The stenosis range is determined based on the contraction locations of multiple consecutive real cavities, and the stenosis assessment result is output based on the stenosis range.

[0079] Specifically, such as Figure 2 As shown, the target vascular image can be CTA image, MRA image, DSA three-dimensional reconstruction image, intravascular ultrasound image, or other medical image data that can reflect the boundary of the vascular cavity. Figure 2 This illustration shows a target vascular image and a vascular cavity model obtained from the target vascular image, used to illustrate the application scenarios of target vascular image input and target vascular cavity model extraction. It does not limit the specific location, image type, or lesion morphology of the target vessel. After acquiring the target vascular image, it undergoes preprocessing, including spatial resolution unification, grayscale normalization, noise suppression, vascular enhancement, and non-vascular region suppression. Spatial resolution unification converts voxel spacing in different directions to a uniform scale; grayscale normalization reduces grayscale differences caused by different acquisition devices or scanning parameters; noise suppression reduces the impact of local speckle, artifacts, and boundary spicules on vascular cavity extraction; and vascular enhancement highlights tubular structures, making the vascular cavity have a clearer boundary difference with surrounding tissues.

[0080] The preprocessed target vascular image is subjected to vascular cavity extraction to obtain a target vascular cavity model. The target vascular cavity model can be a three-dimensional voxel set, a triangular mesh model, or a spatial model composed of boundary points of the vessel wall. If the target vascular image is three-dimensional volumetric data, the vascular cavity region can be extracted using threshold segmentation, region growing, level set evolution, graph cut segmentation, or segmentation methods based on tubular structure response. If the target vascular image already contains vessel segmentation results, the target vascular cavity model can be directly constructed based on those results. The target vascular cavity model is used for subsequent centerline extraction, cross-sectional contour calculation, and radial data analysis of the vessel wall.

[0081] After extracting the target vascular cavity model, such as Figure 3 As shown, the centerline of the target blood vessel is obtained. Figure 3The continuous vascular structure of the target vessel is shown, illustrating that the target vessel's centerline extends along the main vascular trunk and serves as the basis for subsequent sampling location setting, cross-sectional construction, and analysis of the geometric continuity of adjacent cross-sections. The target vessel's centerline can be obtained through vascular skeletonization, center path search, shortest path tracing, or center point extraction based on the cavity distance field. To reduce centerline jitter caused by local noise, the initial centerline can be smoothed and resampled according to the arc length parameter. The target vessel's centerline is denoted as: ,in, This represents the arc length parameter along the centerline of the target blood vessel. Multiple sampling positions are set along the centerline of the target blood vessel. The sampling positions can be set at equal arc length intervals or adaptively set according to the local lumen diameter of the blood vessel. To ensure geometric continuity between adjacent sections, the sampling interval is preferably less than a certain proportion of the average equivalent lumen diameter of the target blood vessel, for example, less than one-half or one-third of the average equivalent lumen diameter.

[0082] For the There are 1 sampling location, and its centerline coordinates are denoted as . A corresponding cross-section is established based on the local tangential direction of the target blood vessel centerline at the sampling location. The local tangential direction can be determined by adjacent centerline sampling points. ,in, For the first The centerline direction at each sampling location. Using the normal vector, establish a cross-sectional plane orthogonal to the local tangent direction. Intersect this cross-sectional plane with the target blood vessel cavity model to obtain the cross-sectional contour corresponding to the current sampling position. If the cross-sectional plane intersects with the target blood vessel cavity model to form multiple closed contours, select the closed contour containing the centerline sampling point or the one closest to the centerline sampling point as the current cross-sectional contour.

[0083] Cross-sectional geometric data is obtained based on the cross-sectional profile. This data includes the cross-sectional area, equivalent cavity diameter, cross-sectional center, cross-sectional perimeter, major axis length, and minor axis length. The cross-sectional area is denoted as... The perimeter of the cross section is denoted as The equivalent lumen diameter can be determined using the equivalent circle diameter calculation method commonly used in existing vascular geometry analysis: ,in, For the first The original actual cavity diameter corresponding to each sampling position. The center of the cross-section can be determined based on the geometric center of the region enclosed by the cross-section profile, denoted as . The major axis length and minor axis length of the cross section can be obtained by ellipse fitting or principal direction projection of the cross section profile, and are denoted as follows: and ,in The degree of eccentricity of the cross-section can also be obtained from its major and minor axes. The degree of eccentricity can be used to subsequently determine the eccentric contraction trend, but it is not limited to participating in all the steps in the claims.

[0084] When acquiring wall radial data, use the cross-section center as the reference point. Starting from the radial origin, along multiple predetermined radial directions within the cross-sectional plane. Sampling is performed, and the multiple preset radial directions can be set at equal angular intervals, such as one direction every 10 degrees, 15 degrees, or 30 degrees. Radial lines are emitted from the center of the cross-section to the cross-sectional contour along each radial direction, and the distance from the center of the cross-section to the cross-sectional contour is obtained, resulting in: ,in, For the first The sampling location at the first sampling position The radial distances of the walls in multiple radial directions constitute the first radial distance. The wall radial data at each sampling location; after arranging the wall radial data from multiple sampling locations according to the sampling order of the target blood vessel centerline, a wall radial data sequence can be formed.

[0085] To mitigate the impact of local cross-sectional oblique cutting, centerline jitter, segmentation burrs, or local contour distortion, a cross-sectional effectiveness sequence is generated based on the geometric continuity relationship between each cross-section and its adjacent cross-sections. For the first... At each sampling location, the continuous product deviation, center offset deviation, and direction deflection deviation are determined respectively.

[0086] Area continuity deviation is used to characterize the first Whether the cross-sectional area of ​​a given section undergoes an abrupt change relative to its adjacent cross-sectional areas. This can be determined by... , No. , No. The cross-sectional areas corresponding to each sampling location are denoted as follows: , , The continuous deviation of the area can be determined based on the continuity relationship between the current cross-sectional area and the adjacent cross-sectional areas. For example, when Significantly greater than or significantly less than by and When the trend of neighborhood area change is represented, the area continuity deviation increases; when When the trend of area change is similar to that of the neighboring area, the continuous deviation of area decreases. The continuous deviation of area can also be determined using the following example: ,in, To prevent the default small amount where the denominator is zero, this formula is used to illustrate how continuous area deviation can be achieved, and does not limit the invention to this form.

[0087] Center offset deviation is used to characterize whether the center of the current section deviates from a continuous trend relative to the centers of adjacent sections. The first... , No. , No. The cross-sectional centers corresponding to each sampling location are respectively denoted as... , , The center offset deviation can be based on Compared to and The offset of the middle position is determined as follows: ,in, For the first The equivalent lumen diameter at each sampling location. By normalizing using the equivalent lumen diameter, the problem of incomparable center offset between blood vessels of different diameters can be avoided.

[0088] Directional deflection deviation is used to characterize whether there is a significant abrupt change between the centerline direction at the current sampling position and the centerline direction at adjacent sampling positions. , No. , No. The centerline directions at each sampling location are denoted as follows: , The directional deflection deviation can be determined based on the change in the angle between adjacent tangent vectors: When adjacent tangent vectors are close in direction, the directional deflection deviation is small; when the centerline direction changes significantly, the directional deflection deviation increases.

[0089] When the area continuity deviation, center offset deviation, and direction deflection deviation are all within their respective first deviation ranges, it indicates that the cross-section has good geometric continuity with adjacent cross-sections. This cross-section is then designated as a high-effective cross-section, and a first-value effective section is generated. When any one of these deviations is within its corresponding second deviation range, and no deviation falls within the third deviation range, it indicates that the cross-section exhibits some degree of geometric discontinuity but can still participate in subsequent corrections. This cross-section is then designated as a medium-effective cross-section, and a second-value effective section is generated. When any one of these deviations is within its corresponding third deviation range, it indicates that the cross-section may be affected by significant oblique cutting, segmentation anomalies, or abrupt changes in centerline direction. This cross-section is then designated as a low-effective cross-section, and a third-value effective section is generated. The first, second, and third-value effective sections are set sequentially from highest to lowest confidence level. For example, the first-value effective section can be set to a high confidence level, the second to a medium confidence level, and the third to a low confidence level. The effective values ​​of multiple sampling locations are arranged according to the sampling order of the target vessel centerline to obtain a cross-section effectiveness sequence.

[0090] After obtaining the cross-sectional validity sequence, the original actual cavity diameter, original actual cross-sectional area, and wall radial data corresponding to each sampling location are subjected to confidence correction. The purpose of confidence correction is to ensure that high-confidence cross-sections retain more of their original measurement results, and that low-confidence cross-sections rely more on the geometric variation trend of their neighborhood, thereby reducing the impact of a single abnormal cross-section on stenosis assessment.

[0091] Specifically, the cross-sectional acceptance weight for each sampling location is determined based on the effective cross-sectional value. High effective cross-sections correspond to higher acceptance weights, medium effective cross-sections to medium acceptance weights, and low effective cross-sections to lower acceptance weights. Subsequently, neighboring cross-sections whose effective cross-sectional values ​​satisfy preset validity conditions are searched within the neighborhood of the current sampling location. These preset validity conditions can be that the effective cross-sectional value belongs to a high or medium effective cross-section, or that the effective cross-sectional value reaches a preset confidence level. Based on the neighboring cross-sections that satisfy the preset validity conditions, the neighborhood extended cavity diameter, neighborhood extended cross-sectional area, and neighborhood extended wall radial data corresponding to the current sampling location are generated.

[0092] The neighboring extended cavity diameter can be determined based on the pre-correction cavity diameter variation trend of adjacent effective cross-sections or the original actual cavity diameter variation trend. For example, when there are effective cross-sections on both sides of the current sampling position, interpolation or trend extension can be performed based on the cavity diameter variation trends of the effective cross-sections on both sides; when there is only one effective cross-section, extrapolation can be performed based on the cavity diameter variation trends of multiple consecutive effective cross-sections on that side. The neighboring extended cross-sectional area and the neighboring extended wall radial data can be generated in a similar manner. Preferably, the neighboring extended wall radial data are extended separately in the same radial direction, that is, for each radial direction, the neighboring extended wall radial data of the current sampling position in that radial direction is determined based on the wall radial distance variation trend of adjacent effective cross-sections in that radial direction.

[0093] Based on the cross-sectional acceptance weights, the original actual cavity diameter at the current sampling location is fused with the neighboring extended cavity diameter to obtain the corrected actual cavity diameter; the original actual cross-sectional area at the current sampling location is fused with the neighboring extended cross-sectional area to obtain the corrected actual cross-sectional area; and the wall radial data at the current sampling location is fused with the neighboring extended wall radial data to obtain the corrected wall radial data. The fusion rules can be as follows: when the current sampling location is a high effective cross-section, the original data is the primary basis, and the neighboring extended data is used for auxiliary correction; when the current sampling location is a medium effective cross-section, the original data and the neighboring extended data jointly participate in the correction; when the current sampling location is a low effective cross-section, the neighboring extended data is the primary basis, reducing the influence of the original data on the correction result. This results in a corrected actual cavity diameter sequence, a corrected actual cross-sectional area sequence, and a corrected wall radial data sequence.

[0094] Subsequently, the region of the vessel to be analyzed is determined along the center line of the target vessel. For example... Figure 4 As shown, the vascular region to be analyzed can be a segment on the center line of the target vascular vessel that is continuously lower than the trend of the adjacent lumen diameter, or it can be a suspected stenosis segment from an external input. Figure 4 The method for determining the vascular region to be analyzed is illustrated. This region extends along the centerline of the target vessel and is used to subsequently construct a reference lumen diameter curve, multidimensional lumen deviation, and stenosis area. The vascular region to be analyzed can be automatically identified by the system or determined by external input. In automatic identification, segments continuously lower than the neighborhood lumen diameter variation trend along the target vessel centerline are identified based on a corrected actual lumen diameter sequence. For example, a local lumen diameter variation trend is first established based on the corrected actual lumen diameter sequence; if the corrected actual lumen diameter at multiple sampling locations within a continuous segment is lower than this local lumen diameter variation trend, and the length of this continuous segment meets a preset length condition, then this continuous segment is determined as the vascular region to be analyzed. For external input, the physician can specify the area to be evaluated on a three-dimensional vascular model, or a suspected stenosis area can be input from the host system.

[0095] After determining the vascular region to be analyzed, proximal and distal reference vascular segments are identified at the proximal and distal ends of this region, respectively. For example... Figure 5 As shown, the proximal reference vessel segment and the distal reference vessel segment are located outside the vessel region to be analyzed, respectively, and are used to provide the reference data required to construct the reference lumen diameter curve, the reference equivalent cross-sectional area sequence, and the reference wall radial sequence. Figure 5 This diagram illustrates the selection of proximal and distal reference vessel segments along the main trunk of the target vessel. The vascular morphology shown is only for illustrating the method of determining the reference vessel segments and does not limit the length, specific location, or vessel part of the reference vessel segments. Specifically, continuous vessel segments are searched along the centerline of the target vessel in both the proximal and distal directions within the vessel interval to be analyzed. The searched continuous vessel segments must meet the following conditions: the effective cross-sectional value meets preset effective conditions, the corrected actual lumen diameter variation meets preset stability conditions, and the corrected actual cross-sectional area variation meets preset stability conditions. Preset stability conditions may include lumen diameter variation below a preset proportion, cross-sectional area variation below a preset proportion, or the lumen diameter variation direction remaining consistent across multiple consecutive sampling locations without abrupt changes. Proximal continuous vessel segments that meet the conditions are identified as proximal reference vessel segments, and distal continuous vessel segments that meet the conditions are identified as distal reference vessel segments. Both proximal and distal reference vessel segments are located outside the vessel interval to be analyzed to avoid using narrowing data within suspected stenosis intervals as normal references.

[0096] Reference lumen diameter curves are constructed based on the corresponding lumen diameter data of the proximal and distal reference vessel segments. Combined with... Figure 5The proximal and distal reference vessel segments shown are used to characterize the reference lumen diameter change trend of the analyzed vessel segment under conditions of no abnormal contraction. Specifically, the lumen diameter data of the proximal and distal reference vessel segments are extracted from the corrected actual lumen diameter sequence to obtain the proximal and distal lumen diameter change trends. A reference lumen diameter curve is constructed along the analyzed vessel segment based on these trends. The reference lumen diameter curve can be generated through proximal-distal trend extension, smooth interpolation, piecewise fitting, or two-end constraint fitting. The reference lumen diameter curve represents the reference lumen diameter change trend of the analyzed vessel segment under conditions of no abnormal contraction. During construction, the direct influence of suspected narrowing data within the analyzed vessel segment on the reference lumen diameter curve should be minimized, ensuring that the reference lumen diameter curve is primarily determined by the lumen diameter data of the proximal and distal reference vessel segments.

[0097] A reference equivalent cross-sectional area sequence is generated based on the reference cavity diameter curve. For each sampling position in the reference cavity diameter curve, the reference equivalent cross-sectional area can be determined using the equivalent circular cross-sectional area calculation method. For example, the first... The reference cavity diameter at each sampling location is denoted as Then the reference equivalent cross-sectional area can be: ,in, For the first The reference equivalent cross-sectional area corresponds to each sampling location. The reference equivalent cross-sectional areas at multiple sampling locations are arranged in the sampling order along the centerline to form a reference equivalent cross-sectional area sequence.

[0098] A reference wall radial sequence is generated based on the reference lumen diameter curve and the corrected radial wall data of the proximal and distal reference vessel segments. Specifically, corrected radial wall data in multiple radial directions are extracted for both the proximal and distal reference vessel segments to determine the variation trend of the radial wall data in each radial direction. For any sampling location within the vessel segment to be analyzed, the overall reference scale at that location is determined based on the reference lumen diameter corresponding to the reference lumen diameter curve. Then, combined with the variation trends of the proximal and distal reference vessel segments in each radial direction, reference wall radial data for that sampling location in each radial direction is generated. The reference wall radial data from multiple sampling locations are arranged in centerline order to form a reference wall radial sequence. In this way, the reference wall radial sequence reflects not only the size of the reference lumen diameter but also the cross-sectional morphology distribution of the normal reference vessel segment, providing a basis for subsequent eccentric contraction assessment.

[0099] Then, a multidimensional cavity deviation is constructed. The first cavity deviation is used to characterize the degree of reduction in the actual cavity diameter relative to the reference cavity diameter curve. For example, the second... The corrected actual cavity diameter at each sampling position is denoted as The reference cavity diameter is denoted as Then the deviation of the first cavity can be determined according to and The difference is determined. The second cavity deviation is used to characterize the degree of area reduction of the corrected actual cross-sectional area relative to the reference equivalent cross-sectional area sequence. For example, the first... The corrected actual cross-sectional area at each sampling location is denoted as... The reference equivalent cross-sectional area is denoted as The deviation of the second cavity can be determined according to... and The difference is determined. The third cavity deviation is used to characterize the degree of wall indentation of the corrected wall radial data relative to the reference wall radial sequence. For example, the third... Each sampling location is in the radial direction The radial distance of the modified wall surface is denoted as The radial distance from the reference wall is denoted as The degree of wall contraction in that direction can be determined according to... The size is determined. The degree of wall indentation in multiple radial directions together forms the deviation of the third cavity.

[0100] For the same sampling location, its authenticity as a cavity contraction location is determined based on the consistency relationship between the multidimensional cavity deviations. Specifically, when both the first and second cavity deviations meet the preset contraction conditions, and the difference between the first and second cavity deviations is within a preset consistency range, it indicates that the sampling location simultaneously exhibits both cavity diameter and cross-sectional area reduction, and the degree of change is consistent. This sampling location is then determined to satisfy the preset consistency relationship. If the first cavity deviation meets the preset contraction conditions but the second cavity deviation does not, or the second cavity deviation meets the preset contraction conditions but the first cavity deviation does not, or the difference between the two exceeds the preset consistency range, it indicates that the sampling location may exhibit cross-sectional anomalies, wall contour disturbances, or localized pseudo-contraction. This sampling location is then determined to not satisfy the preset consistency relationship.

[0101] For sampling locations that do not meet the preset consistency relationship, their corrected actual cavity diameter, corrected actual cross-sectional area, and corrected wall radial data are updated. The update can be based on the effective cross-sectional value of the sampling location, the multidimensional cavity deviation of adjacent sampling locations, neighborhood extension data, and data from adjacent sampling locations that meet the consistency relationship. After the update, the first cavity deviation, second cavity deviation, and third cavity deviation of the sampling location are recalculated, and the satisfaction of the preset consistency relationship is reassessed. This process forms a closed-loop data update, which can reduce misjudgments caused by errors in a single cross-section.

[0102] For sampling locations that satisfy a preset consistency relationship, they are determined as the actual cavity contraction locations. Then, the contraction pattern of the actual cavity contraction location is determined based on the third cavity deviation. When the third cavity deviation corresponding to the actual cavity contraction location exhibits unilateral wall contraction in the relative radial direction, the contraction pattern of this actual cavity contraction location is determined to be an eccentric contraction pattern. Specifically, this can be: in a certain radial direction, the degree of wall contraction meets a preset wall contraction condition, while the radial direction opposite to this radial direction does not meet the preset wall contraction condition; then, unilateral wall contraction is considered to exist at this location. When the third cavity deviation corresponding to the actual cavity contraction location exhibits wall contraction in multiple radial directions, its contraction pattern is determined to be a centripetal contraction pattern. Specifically, this can be: the degree of wall contraction in at least two relative radial directions meets the preset wall contraction condition, or the degree of wall contraction in multiple radial directions as a whole meets the preset centripetal contraction condition; then, centripetal contraction is considered to exist at this location.

[0103] Finally, the narrowing range of action was determined based on the contraction locations of multiple consecutive real cavities. Combined with... Figure 4 The illustrated vascular region to be analyzed is defined as a stenosis zone if multiple real lumen contraction positions within this region are continuously distributed along the centerline of the target vessel, and the continuous length meets a preset condition. If a small number of sampling positions that do not meet the consistency relationship exist between two real lumen contraction positions, but these sampling positions are continuous with the contraction trends on both sides after updating, they can also be included in the same stenosis zone. The vascular stenosis assessment results are output based on the stenosis zone. These results may include the starting and ending positions of the stenosis zone, the length of the stenosis zone, the minimum corrected actual lumen diameter, the minimum corrected actual cross-sectional area, the maximum deviation of the first lumen, the maximum deviation of the second lumen, the distribution of the deviation of the third lumen, the contraction morphology, and the degree of stenosis. The output format can be a numerical report, a structured table, a 3D vascular model annotation result, or a data file recognizable by the imaging workstation.

[0104] Through the complete processing flow described above, this invention first evaluates the reliability of cross-sectional data using a cross-sectional validity sequence, and then performs reliability correction on the original actual lumen diameter, original actual cross-sectional area, and radial wall data based on the cross-sectional validity sequence. Subsequently, it constructs a reference lumen diameter curve, a reference equivalent cross-sectional area sequence, and a reference radial wall sequence based on the proximal and distal reference vessel segments. Then, it identifies the true lumen contraction location through the consistency relationship between the first, second, and third lumen deviations, and further determines the contraction morphology and stenosis range. This reduces the impact of vessel curvature, oblique cross-sections, abnormal local segmentation, reference baseline deviation, and eccentric pseudo-contraction on the vascular stenosis assessment results, improving the stability of stenosis range identification and stenosis degree assessment.

[0105] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for assessing vascular stenosis based on the analysis of vascular geometric properties, characterized in that, include: Acquire images of the target blood vessel and extract a model of the target blood vessel cavity; Extract the centerline of the target blood vessel, set multiple sampling positions along the centerline of the target blood vessel, establish corresponding cross sections at each sampling position, and obtain the cross section geometric data and wall radial data of each cross section; Based on the geometric continuity between each cross section and adjacent cross sections, a cross section validity sequence distributed along the centerline of the target blood vessel is generated. Based on the cross section validity sequence, the credibility of the original actual lumen diameter, original actual cross-sectional area and wall radial data corresponding to each sampling position is corrected to obtain the corrected actual lumen diameter sequence, corrected actual cross-sectional area sequence and corrected wall radial data sequence. The vascular interval to be analyzed is determined on the centerline of the target vascular segment. Based on the cross-sectional validity sequence and the corrected actual lumen diameter sequence, the proximal reference vascular segment and the distal reference vascular segment are determined at the proximal and distal ends of the vascular interval to be analyzed, respectively. The reference lumen diameter curve is constructed based on the corresponding lumen diameter data of the proximal and distal reference vascular segments. The reference equivalent cross-sectional area sequence is generated based on the reference lumen diameter curve. The reference wall radial sequence is generated based on the reference lumen diameter curve and the corrected wall radial data of the proximal and distal reference vascular segments. Based on correcting the differences between the actual cavity diameter sequence and the reference cavity diameter curve, correcting the differences between the actual cross-sectional area sequence and the reference equivalent cross-sectional area sequence, and correcting the differences between the wall radial data sequence and the reference wall radial data sequence, a multidimensional cavity deviation is constructed, including the first cavity deviation, the second cavity deviation, and the third cavity deviation. Based on the consistency relationship between the multidimensional cavity deviations, the corrected actual cavity diameter, corrected actual cross-sectional area, and corrected wall radial data of the sampling positions that do not meet the preset consistency relationship are updated. The actual cavity contraction position and contraction shape are determined based on the multidimensional cavity deviations that meet the preset consistency relationship. The stenosis range is determined based on the contraction locations of multiple consecutive real cavities, and the stenosis assessment result is output based on the stenosis range.

2. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Establish corresponding cross sections at each sampling location, and obtain the cross section geometry data and wall radial data for each cross section, including: Based on the local tangential direction of the target blood vessel centerline at the current sampling position, establish a cross-sectional plane orthogonal to the local tangential direction; Intersecting the cross-sectional plane with the target blood vessel cavity model yields the cross-sectional profile corresponding to the current sampling position; Based on the cross-sectional profile, the cross-sectional area, equivalent cavity diameter, cross-sectional center, cross-sectional perimeter, cross-sectional major axis length, and cross-sectional minor axis length are obtained as cross-sectional geometric data; Using the center of the cross section as the radial starting point, the distance from the center of the cross section to the cross section profile is obtained along multiple preset radial directions to obtain the radial data of the wall surface.

3. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Based on the geometric continuity between each section and its adjacent sections, a sequence of effective sections distributed along the centerline of the target blood vessel is generated, including: For the cross section corresponding to the i-th sampling position, calculate the area continuity deviation between the cross section area and the cross section area of ​​the adjacent cross section; Calculate the center offset deviation between the center of this section and the center of the adjacent section; Calculate the directional deviation between the centerline direction at the sampling location and the centerline direction at the adjacent sampling location; Based on area continuity deviation, center offset deviation, and direction deflection deviation, the effective value of the cross section corresponding to the i-th sampling position is generated; The effective values ​​of the cross sections corresponding to multiple sampling locations are arranged in the sampling order of the target blood vessel centerline to obtain the cross section effectiveness sequence.

4. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 3, characterized in that, Based on area continuity error, center offset error, and direction deflection error, the effective cross-sectional value corresponding to the i-th sampling position is generated, including: The area continuity deviation, center offset deviation, and direction deflection deviation are compared with their respective preset deviation ranges. When the area continuity deviation, center offset deviation, and direction deflection deviation are all within the corresponding first deviation range, the cross section corresponding to the i-th sampling position is determined as the high effective cross section, and the first effective value of the cross section is generated. When any one of the area continuous deviation, center offset deviation, and direction deflection deviation is within the corresponding second deviation range, and there is no deviation within the corresponding third deviation range, the cross section corresponding to the i-th sampling position is determined as the effective cross section, and the effective value of the second cross section is generated. When any one of the area continuity deviation, center offset deviation, and direction deflection deviation falls within the corresponding third deviation range, the cross section corresponding to the i-th sampling position is determined as the low effective cross section, and the effective value of the third cross section is generated; wherein, the first deviation range, the second deviation range, and the third deviation range are set in order of increasing deviation degree, and the effective value of the first cross section, the effective value of the second cross section, and the effective value of the third cross section are set in order of decreasing cross section confidence degree; The effective values ​​of the cross sections corresponding to multiple sampling locations are arranged in the sampling order of the target blood vessel centerline to obtain the cross section effectiveness sequence.

5. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Based on the cross-sectional validity sequence, the reliability of the original actual cavity diameter, original actual cross-sectional area, and wall radial data corresponding to each sampling location is corrected, including: The cross-sectional acceptance weight of each sampling location is determined based on the effective value of the cross-section at each sampling location. Based on the cross-sectional effective values ​​of adjacent sampling positions that meet the preset effective conditions, generate the neighboring extended cavity diameter, neighboring extended cross-sectional area, and neighboring extended wall radial data corresponding to the current sampling position; Based on the cross-sectional acceptance weight, the original actual cavity diameter at the current sampling position is fused with the neighboring extended cavity diameter to obtain the corrected actual cavity diameter; The original actual cross-sectional area at the current sampling location is fused with the extended cross-sectional area of ​​the neighboring area to obtain the corrected actual cross-sectional area; The wall radial data at the current sampling location is fused with the neighboring extended wall radial data to obtain the corrected wall radial data.

6. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Determine the region of the vessel to be analyzed along the center line of the target vessel, including: Based on the corrected actual lumen diameter sequence, segments along the centerline of the target blood vessel that are continuously lower than the changing trend of the lumen diameter in the neighborhood are identified; The segments whose diameter changes continuously below the trend of neighboring vessels are identified as the vascular intervals to be analyzed.

7. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Proximal and distal reference vessel segments are identified at the proximal and distal ends of the vascular region to be analyzed, respectively, including: Search for continuous vascular segments in both the proximal and distal directions of the vascular region to be analyzed; Continuous vascular segments whose effective cross-sectional values ​​meet preset effective conditions, whose corrected actual lumen diameter changes meet preset stable conditions, and whose corrected actual cross-sectional area changes meet preset stable conditions are respectively identified as proximal reference vascular segments and distal reference vascular segments; wherein, both proximal and distal reference vascular segments are located outside the vascular interval to be analyzed.

8. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, Constructing the reference cavity diameter curve, the reference equivalent cross-sectional area sequence, and the reference wall radial sequence includes: Based on the lumen diameter variation trends of the proximal and distal reference vessel segments in the corrected actual lumen diameter sequence, a reference lumen diameter curve is constructed along the vessel interval to be analyzed. Based on the reference cavity diameter at the corresponding position of the reference cavity diameter curve, generate a reference equivalent cross-sectional area sequence; Based on the changing trends of the corrected wall radial data in multiple radial directions of the proximal and distal reference vessel segments, and combined with the reference lumen diameter at the corresponding position of the reference lumen diameter curve, reference wall radial data at each sampling position within the vessel segment to be analyzed are generated, forming a reference wall radial sequence.

9. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, The first cavity deviation is used to characterize the degree of reduction in cavity diameter relative to the reference cavity diameter curve; The second cavity deviation is used to characterize the degree of area reduction of the corrected actual cross-sectional area relative to the reference equivalent cross-sectional area sequence; The third cavity deviation is used to characterize the degree of wall indentation of the corrected wall radial data relative to the reference wall radial sequence; When the deviation of the first cavity and the deviation of the second cavity at the same sampling position both meet the preset contraction condition, and the difference between the deviation of the first cavity and the deviation of the second cavity is within the preset consistency range, it is determined that the sampling position meets the preset consistency relationship.

10. The method for assessing vascular stenosis based on vascular geometric property analysis according to claim 1, characterized in that, The actual cavity contraction location and contraction pattern are determined based on the multidimensional cavity deviation that satisfies a preset consistency relationship, including: When the sampling location meets the preset consistency relationship, the sampling location is determined as the actual cavity contraction location; When the deviation of the third cavity corresponding to the actual cavity contraction position is manifested as unilateral wall retraction in the relative radial direction, the contraction pattern of the actual cavity contraction position is determined to be an eccentric contraction pattern. When the deviation of the third cavity corresponding to the actual cavity contraction position is manifested as wall inward contraction in multiple radial directions, the contraction pattern of the actual cavity contraction position is determined to be the centripetal contraction pattern. The contraction locations of multiple real cavities continuously distributed along the centerline of the target blood vessel are determined as the stenosis range. Based on the minimum corrected actual lumen diameter, minimum corrected actual cross-sectional area, stenosis range length, and lumen deviation distribution within the stenosis range, the vascular stenosis assessment result is output.

Images