A DSA coronary artery stenosis analysis method based on matching degree measurement

Abstract

This paper proposes a method for analyzing coronary artery stenosis using DSA (digital subtraction angiography) based on matching measurement. This method combines image preprocessing, catheter calibration, vessel segmentation, and centerline extraction on DSA images with an improved MoM (Measure of Match) model to accurately measure vessel diameter based on matching and calculate the stenosis rate of the stenotic area. This method also features an automatic correction function, enabling automated correction of the stenotic area based on the reference diameter of a normal vessel after stenosis detection, thereby improving detection accuracy and stability. This method is suitable for rapid and accurate analysis of coronary artery stenosis, effectively reducing manual operation errors and demonstrating a high degree of automation.

Description

Technical Field

[0001] The present invention relates to the field of medical image analysis, and in particular to a method for analyzing coronary artery stenosis using digital subtraction angiography (DSA) based on matching degree measurement, which is suitable for automated vascular stenosis detection and correction, thereby improving detection accuracy and efficiency. Background Art

[0002] Coronary artery disease is one of the leading causes of death and disability worldwide. Coronary artery stenosis, especially severe stenosis, can impede blood flow and lead to myocardial ischemia, which can cause serious consequences such as angina and myocardial infarction. Therefore, accurately detecting the location and extent of coronary artery stenosis is of great clinical significance for the diagnosis and treatment of cardiovascular disease.

[0003] Digital subtraction angiography (DSA), the gold standard for detecting cardiovascular stenosis, provides high-resolution images of vascular structures. However, traditional DSA analysis relies on the physician's experience and subjective judgment, which is subject to limitations and errors. Manual analysis is particularly time-consuming and labor-intensive when processing large amounts of imaging data, and the results are susceptible to subjective influences from the physician's operation and judgment. In recent years, with the rapid development of medical image processing technology, computer-based automated vascular analysis methods have been widely studied to improve detection accuracy and reduce human error.

[0004] Although the automated DSA analysis methods in the existing technology can help improve efficiency to a certain extent, they generally have some significant shortcomings. Specifically, these shortcomings include the accuracy of vascular segmentation in complex backgrounds, that is, the vascular segmentation algorithm is easily affected by noise, resulting in inaccurate segmentation of vascular contours, especially in low-contrast images, where accurate segmentation of vascular edges is particularly difficult; stenosis detection errors, especially when the stenosis area is discontinuous or only slightly stenotic, the existing methods often miss detection or misjudge; and the lack of effective automatic correction function. Even if the stenosis area is detected, its automatic correction function is limited and still relies on manual operation for further adjustment, making it impossible to achieve high-precision automatic correction. In addition, in the detection of coronary artery stenosis, the inaccuracy of the catheter calibration factor may also lead to deviations in the calculation results of the degree of stenosis, thereby affecting subsequent clinical judgments. Summary of the Invention

[0005] The present invention addresses the shortcomings of existing coronary artery stenosis detection technologies by proposing a DSA coronary artery stenosis analysis method based on matching degree measurement. By combining automated image processing, precise stenosis measurement, and intelligent correction, this method provides an efficient, accurate, and automated coronary artery stenosis analysis system. This invention not only improves the accuracy of image analysis and reduces subjective errors, but also significantly enhances detection efficiency, making the stenosis detection and correction process more automated and intelligent.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows: a DSA coronary artery stenosis analysis method based on matching degree measurement, the specific steps are as follows:

[0007] Step 1: DSA image reading and preprocessing.

[0008] First, the system reads DSA images from the image database and supports a variety of formats, especially DICOM sequence image processing. To ensure the image processing effect, the system provides the function of adjusting the window and window level to make the image contrast and brightness meet the needs of the application user.

[0009] (1) Calculation of window width and window position: The window width (Width) is defined as the difference between the maximum and minimum CT values ​​in the image, which is used to determine the brightness range of the image. The calculation formula is:

[0010] Width=CT max -CT min

[0011] The window level is defined as the average of the maximum and minimum CT values ​​and is used to determine the median brightness of the image. The calculation formula is:

[0012]

[0013] (2) Image normalization: After calculating the window width and window position, the system will normalize the image. The image brightness will be adjusted according to the calculated window width and window position, and the CT value of the image will be mapped to the range of [0, 255] to facilitate subsequent processing and analysis.

[0014] Step 2: Catheter calibration. Accurately measuring vessel diameters in DSA images requires precise catheter calibration. Catheter calibration aims to calculate a calibration factor by marking and measuring the catheter area, which is used to convert the pixel size in the image to the actual physical size. The steps for catheter calibration are as follows:

[0015] (1) Catheter area marking: First, the catheter's marker points are determined in the DSA image. The catheter is often an important reference object in the image. The system marks the starting and ending points of the catheter through automated or user-interactive methods. Based on these marker points, the system can extract the catheter's outline and determine the catheter's edge area.

[0016] (2) Calculate the catheter diameter: Next, the system calculates the catheter diameter by extracting the contour of the catheter area. This diameter is calculated based on the pixel value of the catheter in the image and may vary in actual application due to differences in image resolution and viewing angle.

[0017] (3) Calculation of calibration factor: In order to convert the pixel size in the image to the actual physical size, the calibration factor of the catheter needs to be calculated. The calibration factor κ is calculated by the following formula:

[0018]

[0019] Among them, D actual is the actual diameter of the catheter (usually in millimeters), and D pixel is the diameter of the catheter in the image (in pixels). This calibration factor can be used to convert the vessel diameter value in the image to its actual physical size.

[0020] (4) Applying the calibration factor: After the catheter is calibrated, the calibration factor is applied to the subsequent vessel diameter measurement. This factor converts the pixel-level dimensions of the vessel to its actual size, ensuring the accuracy of the stenosis calculation.

[0021] The catheter calibration step is a key link in the entire vascular analysis process. It can effectively reduce errors caused by factors such as image scale, viewing angle and resolution, thereby providing an accurate foundation for subsequent vascular analysis.

[0022] Step 3: Vessel Segmentation and Centerline Extraction: The system automatically detects the user's manually marked points on the vessel, identifies the stenosis area, and performs segmentation and centerline extraction on the vessels within the area.

[0023] The blood vessel segmentation in the present invention is achieved through adaptive thresholding technology, which dynamically adjusts the threshold based on the change of pixel intensity in the local area to ensure that the blood vessel area can be accurately segmented under different contrast and lighting conditions.

[0024] Adaptive thresholding:

[0025] For each pixel, its neighborhood is defined as a window of size N×N, which contains the pixel information of the pixel and its neighborhood. The average value of the pixel value in the area is μ local Calculated by the following formula:

[0026]

[0027] Among them, I(i,j) represents the grayscale value of the pixel in the window.

[0028] Next, the threshold T(x,y) of the pixel is determined by calculating the average value within the local area and subtracting a constant C:

[0029] T(x,y)=μ local (x,y)-C

[0030] Thus, for each pixel in the image, the local threshold T(x,y) is calculated based on the pixel values ​​within the neighborhood, rather than a global fixed threshold. This mechanism enables the segmentation algorithm to adapt to changing contrast conditions, ensuring accurate identification of blood vessel regions even in complex backgrounds or uneven lighting. After segmentation, the binary image B(x,y) is defined as follows:

[0031]

[0032] Morphological closing operation:

[0033] After obtaining the initial binary image, the image is further optimized using the morphological closing operation. The closing operation is composed of the image dilation operation D(B) and the erosion operation E(D(B)), and its mathematical definition is as follows:

[0034] Closing operation (B) = E (D (B))

[0035] D(B) represents the dilation operation of the binary image, which expands the edges of the blood vessels and fills the holes in the segmentation image. E(D(B)) represents the erosion operation after the dilation, which shrinks the expanded portion to preserve the shape and continuity of the blood vessels. This step ensures the integrity of the blood vessel region, especially for broken or small blood vessel segments. The closing operation can restore them to a more coherent shape and reduce the interference of background noise.

[0036] Centerline extraction:

[0037] After vessel segmentation, this invention employs a skeleton extraction technique to ensure that the centerline of the vascular structure can be extracted from the binary vascular image. This technique progressively reduces the pixels in the vascular region while preserving its topological structure. The resulting centerline has a single-pixel width and fully preserves the connectivity and morphological characteristics of the vessel. This process ensures that the path of the main vessels in complex vascular networks is not disrupted, laying the foundation for subsequent analysis.

[0038] Centerline pruning:

[0039] After initially extracting the centerline of the blood vessel, further centerline pruning is required to remove redundant noise segments and ensure the connectivity and accuracy of the centerline. The pruning process specifically includes the following steps:

[0040] (1) Determination of the starting and ending points: To ensure alignment between the centerline and the vessel segmentation markers, the present invention uses a method based on the Euclidean distance (L2 norm) to determine the starting and ending points of the centerline. The distances between the vessel segmentation markers (i.e., the first and last markers) and each point on the centerline are calculated, and the centerline point with the smallest distance is selected as the starting and ending points. The calculation formula for the L2 distance d(x, y) is:

[0041]

[0042] Where (x1, y1) is the coordinate of the marker point, and (x2, y2) is the coordinate of the centerline point. Through this distance calculation, we can accurately find the centerline node closest to the marker point.

[0043] (2) Depth-first search (DFS) path tracing: After determining the starting and ending points, a depth-first search (DFS) algorithm is used to trace the path along the centerline. Starting from the starting point, DFS recursively traverses all connected nodes on the centerline, searching along the vascular path until the ending point is found. DFS ensures that the selected path is a connected path from the starting point to the ending point.

[0044] (3) Path Verification and Pruning: Once DFS finds a path to the endpoint, the system retains that path and prunes invalid centerline segments that are not included in the main path. This step effectively removes excess noise and unnecessary branches, ensuring that the resulting centerline is both streamlined and consistent with the actual vascular structure.

[0045] In this way, the present invention can accurately extract a complete and coherent blood vessel centerline, providing a reliable basis for subsequent blood vessel diameter measurement and stenosis analysis.

[0046] Step 4: Vessel Diameter Measurement. After centerline extraction, the system interpolates the vessel contours and fits the vessel edges using a B-spline curve to ensure accurate reconstruction of the vessel morphology. To further improve the accuracy of diameter calculation, especially in curved or narrow areas, the present invention proposes and utilizes an improved MoM (Measure of Match) model.

[0047] (1) Determination of blood vessel direction: In blood vessel contour fitting, the system first determines the direction of the blood vessel through curve fitting and Hessian matrix optimization. Assuming point P is the current center point, the slope θ of the local fitting line segment at point P is weighted and combined. p1and the eigenvector θ corresponding to the smaller eigenvalue of the Hessian matrix p2 As the blood vessel direction θ p , defined as follows:

[0048] θ p =αθ p1 +(1-α)θ p2

[0049] (2) Calculation of blood vessel diameter:

[0050] The system calculates the vessel diameter based on the distance d between the left and right boundary points P1 and P2. The improved MoM model introduces an optimization of the matching degree of bilateral boundary points, making the boundary features on both sides clearer. The diameter calculation formula is:

[0051] Blood vessel diameter = d l +d r

[0052] Among them, d l and d r They represent the distances from the center point P to the left and right boundary points P1 and P2 respectively, and the diameter of the blood vessel is obtained by summing them.

[0053] (3) Edge matching measurement:

[0054] The improved MoM algorithm maximizes the edge matching by matching the actual edge of the blood vessel with the fitted curve. To improve the accuracy of the calculation, the algorithm calculates the matching degree of the left and right sides separately:

[0055]

[0056] Where m is the maximum search distance, f1(t+i) and f2(t+i) are the fitted left and right curves, respectively, and g1(t+i) and g2(t+i) are the left and right vessel edge points. By calculating the deviation between the fitted curve and the actual edge, the improved MoM algorithm maximizes the matching between the left and right vessel edges, further optimizing the accuracy of vessel diameters.

[0057] (4) Determination of blood vessel diameter:

[0058] Finally, the system determines the optimal boundary point positions P1 and P2 on the left and right sides by maximizing the MoM values ​​(matching degree) on the left and right sides:

[0059]

[0060] The method proposed by the present invention can effectively improve the accuracy of diameter measurement, especially when dealing with curved or narrow blood vessel segments.

[0061] Step 5: Detection and measurement of stenosis. By comparing the vessel diameters of the stenosis area and the normal area, the system automatically identifies the stenosis area and calculates the stenosis rate based on the minimum diameter of the stenosis area and the reference diameter. The specific calculation formula is as follows:

[0062]

[0063] Among them, %D is the diameter stenosis rate, %A is the area stenosis rate, and D Lesion (Unit: mm) is the minimum diameter of the area with vascular stenosis rate ≥5%, D Ref (Unit: mm) is the average diameter of the healthy vessel segments adjacent to the stenotic area. When the stenosis rate exceeds the preset threshold of 5%, the system automatically marks the area for further review or subsequent processing.

[0064] Step 6: Automatic Stenosis Correction. After stenosis detection, the system automatically corrects the stenosis area. Using the vessel centerline and reference diameter as references, the system uses a B-spline curve to generate a reference contour of a normal vessel and automatically corrects the abnormal diameter in the stenosis area. This automatic correction function not only improves the accuracy of stenosis detection but also reduces the workload of manual adjustments. The corrected vessel contour is smooth and conforms to the vessel's natural morphology.

[0065] Compared with the prior art, the advantages of the present invention are as follows: (1) High precision and automation. Through catheter calibration, improved MoM model and B-spline curve interpolation, the present invention can provide extremely high stenosis detection and diameter measurement accuracy, and realize fully automatic stenosis area correction, which greatly reduces the subjective error of manual operation. (2) Strong adaptability. The adaptive threshold segmentation algorithm of the present invention can adapt to DSA images with different contrast and image quality, especially in images with low contrast and complex background, it still maintains high segmentation accuracy. (3) Automatic correction function. The automatic stenosis correction function provided by the present invention can not only effectively process the detected stenosis area, but also ensure the naturalness and consistency of the vascular morphology through automatic correction based on the reference contour of the normal blood vessel, reducing subsequent manual adjustment. (4) User-friendliness. The system has a friendly user interface, supports manual adjustment and dynamic display of results, can meet the personalized processing needs of clinicians for different cases, and improves the convenience of clinical application. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] Figure 1 The system interface, function introduction and flow chart of the present invention;

[0067] Figure 2 Schematic diagram for reading the original image;

[0068] Figure 3 This is a schematic diagram of catheter calibration;

[0069] Figure 4 To determine the narrow area marking points, calculate the segmentation map and centerline diagram;

[0070] Figure 5 Diameter measurement diagram;

[0071] Figure 6 Schematic diagram of stenosis measurement;

[0072] Figure 7 Schematic diagram of automatic stenosis correction. DETAILED DESCRIPTION

[0073] The present invention is further illustrated below with reference to specific examples. It should be understood that these examples are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0074] Example: Figure 1 As shown, the present invention provides a DSA coronary artery stenosis analysis method based on matching measurement, and the final test results are as follows Figure 6 The specific steps are as follows:

[0075] Step 1: Read the original image. First, read the multi-frame DSA image sequence in DICOM format, and adjust the window width and window position by sliding the right wheel of the system so that the contrast and brightness of the blood vessels meet the analysis requirements. The adjusted image will be used for further blood vessel segmentation and feature extraction. Figure 2 shown.

[0076] Step 2: Catheter calibration. Mark the starting and ending points of the catheter in the image, calculate the ratio of the actual diameter of the catheter to the pixel value in the image, and generate a calibration factor for subsequent vessel diameter measurement to convert pixel units to physical units. The calculation formula for the catheter calibration factor is:

[0077]

[0078] Among them, D actual is the actual diameter of the catheter, and D pixel is the diameter of the catheter in the image. Figure 3 shown.

[0079] Step 3, determine the marker points of the stenosis area, calculate the segmentation map and center line. Through image analysis technology, combined with the morphological characteristics of the blood vessel contour, the position of the marker points in the stenosis area is automatically calculated to accurately cover the stenosis area. The adaptive threshold segmentation method is used to binarize the blood vessel area, and the local threshold is dynamically adjusted to ensure accurate segmentation of blood vessels under complex backgrounds. Then, the morphological closing operation is used to remove small holes and noise, and the continuity of the blood vessels is retained. Subsequently, the skeleton extraction technology is used to generate the blood vessel centerline, and the depth-first search (DFS) algorithm is used for pruning to ensure the accuracy and continuity of the centerline. The effect is as follows Figure 4 shown.

[0080] Step 4: Blood vessel diameter measurement. After completing step 3, the blood vessel contour is calculated using the improved MoM (Measure of Match) model. By combining B-spline curve interpolation technology, the system can accurately measure the blood vessel diameter at different locations, especially when dealing with curved and narrow blood vessels. By optimizing the left and right boundary point matching measurement, the blood vessel diameter calculation is more accurate. The effect is as follows: Figure 5 shown.

[0081] Step 5: Measure the stenosis area. Based on the vessel diameter obtained in step 4, the stenosis rate is automatically calculated by comparing the diameter of the stenosis area with the diameter of the healthy vessel area. The stenosis rate calculation formula is:

[0082]

[0083] Among them, %D is the diameter stenosis rate, %A is the area stenosis rate, and D Lesion (Unit: mm) is the minimum diameter of the area with vascular stenosis rate ≥5%, D Ref (Unit: mm) is the average diameter of the healthy blood vessel segment adjacent to the stenosis area. When the stenosis rate exceeds 5%, the system automatically marks the area. Figure 6 shown.

[0084] Step 6: Automatic stenosis correction. Based on the stenosis measurement results of step 5, the system automatically refers to the diameter of the healthy blood vessel, uses the B-spline curve to generate the reference contour of the stenosis area, and corrects the diameter of the abnormal stenosis area. This correction method ensures that the stenosis area returns to normal vascular morphology, reduces manual adjustment operations, and improves the overall degree of automation. The effect is as follows Figure 7 shown.

[0085] Effect evaluation:

[0086] The present invention discloses a DSA coronary artery stenosis analysis method based on matching degree measurement. The test results of this method are as follows Figure 6As shown in the accompanying drawings, the present invention is integrated into a software system. The testing process is as follows: click the "Function" button in the upper left corner of the system, find the "Read Original Image" button to import the DSA image into the system, determine the catheter marker points, click the "Catheter Calibration" button to obtain the calibration factor, then click the "Calculate Segmentation Map" button based on the stenosis area marker points to obtain the vessel segmentation map and centerline, then click the "Calculate Diameter" button to calculate the vessel segment diameter, then click the "Stenosis Measurement" button to calculate the vessel stenosis rate, and finally click the "Automatic Stenosis Correction" button to restore the healthy vessel segment reference. Test results show that this method can quickly and accurately detect coronary artery stenosis areas, providing doctors with an efficient and reliable analysis tool for the diagnosis and treatment of coronary artery disease.

[0087] The above description is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person familiar with the technology can understand and think of any changes or replacements within the technical scope disclosed by the present invention, which should be included in the scope of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

Claims

1. A DSA coronary artery stenosis analysis method based on matching degree measurement, characterized in that: The method comprises the following steps: a) Image reading and preprocessing: Read DSA images and adjust image window width and window position to optimize image display quality; b) Catheter calibration: Select the catheter markers in the image in the software and calculate the catheter calibration factor for subsequent accurate measurement of vessel diameter and stenosis. The catheter calibration factor is the ratio of the actual catheter width (mm) to the measured pixel value (pixel); c) Blood vessel segmentation: The blood vessel region is binarized based on the adaptive threshold segmentation method to generate a segmentation map of the blood vessel region; d) Centerline extraction: Extract the vascular skeleton from the segmentation image and remove noise and disconnected areas through pruning to ensure the continuity of the centerline; e) Vascular diameter measurement: The improved MoM model is used to calculate the vascular contour, and the vascular diameter at different locations is accurately measured through B-spline curve interpolation; f) Stenosis area measurement: By comparing the diameters of the stenotic area and the normal vascular area, the vascular stenosis rate is calculated. The stenosis rate is calculated using the following methods: Among them, %D is the diameter stenosis rate, %A is the area stenosis rate, and D Lesion (Unit: mm) is the minimum diameter of the area with vascular stenosis rate ≥5%, D Ref (Unit: mm) is D Lesion Normal blood vessel diameter reference value, g) Automatic stenosis correction: Based on the reference diameter and the vessel centerline, the reference contour of the normal vessel is automatically generated, and the diameter of the stenosis area is corrected through B-spline curve interpolation.

2. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 1, characterized in that: ； Step 1: DSA image reading and preprocessing. First, the system reads DSA images from the image database and supports multiple formats, especially the processing of DICOM sequence images. In order to ensure the image processing effect, the system provides the function of adjusting the window and window level to make the contrast and brightness of the image meet the needs of the application user. (1) Calculation of window width and window position: The window width (Width) is defined as the difference between the maximum and minimum CT values ​​in the image, which is used to determine the brightness range of the image. The calculation formula is: Width=CT max -CT min The window level is defined as the average of the maximum and minimum CT values ​​and is used to determine the median brightness of the image. The calculation formula is: (2) Image standardization: After calculating the window width and window position, the system will standardize the image. The image brightness will be adjusted according to the calculated window width and window position, and the CT value of the image will be mapped to the range of [0, 255] to facilitate subsequent processing and analysis.

3. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 1, characterized in that: The catheter calibration steps include: a) selecting marker points in the catheter area and binarizing the catheter area through adaptive threshold segmentation; b) extracting the catheter centerline and calculating the catheter calibration factor; c) The catheter calibration factor is calculated as the ratio of the actual catheter width (mm) to the measured pixel value (pixels).

4. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 1, characterized in that: The adaptive threshold segmentation method in the blood vessel segmentation step dynamically adjusts the threshold according to the local contrast of the image to ensure the accuracy of blood vessel segmentation, especially in low-contrast areas; Adaptive thresholding: For each pixel, its neighborhood is defined as a window of size N×N, which contains the pixel information of the pixel and its neighborhood. The average value of the pixel value in the area is μ local Calculated by the following formula: Among them, I(i,j) represents the grayscale value of the pixel in the window, Next, the threshold T(x,y) of the pixel is determined by calculating the average value within the local area and subtracting a constant C: T(x,y)=μ local (x,y)-C In this way, for each pixel in the image, its local threshold T(x,y) is calculated based on the pixel values ​​in the neighborhood, rather than a global fixed threshold. This mechanism enables the segmentation algorithm to adapt to changing contrast conditions, ensuring that blood vessel areas can still be accurately identified in the case of complex backgrounds or uneven lighting. After segmentation, the binary image B(x,y) is defined as follows:

5. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 1, characterized in that: The centerline extraction step comprises: a) Using a breadth-first search algorithm, starting from the marked point, traverse the connectivity of the vascular region; b) Remove noise and disconnected small segments using the skeleton pruning algorithm to ensure the continuity of the centerline; The pruning process specifically includes the following steps: (1) Determination of the starting and ending points: The starting and ending points of the centerline are determined using a method based on Euclidean distance (L2 norm). The distances between the vessel segmentation markers (i.e., the first and last markers) and each point on the centerline are calculated. The centerline points with the smallest distance are selected as the starting and ending points. The L2 distance calculation formula is: Among them, (x1, y1) is the coordinate of the mark point, and (x2, y2) is the coordinate of the center line point. Through this distance calculation, the center line node closest to the mark point is accurately found. (2) Depth-first search (DFS) path tracking: After determining the starting point and the end point, the depth-first search (DFS) algorithm is used to track the path on the center line. DFS starts from the starting point and recursively traverses all connected nodes on the center line, searching along the blood vessel path until the end point is found. DFS ensures that the selected path is a connected path from the starting point to the end point. (3) Path verification and pruning: Once DFS finds a path to the end point, the system will retain the path and prune the invalid centerline segments that are not included in the main path. This step can effectively remove excess noise and unnecessary branches, ensuring that the final centerline is both streamlined and consistent with the actual vascular structure.

6. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 1, characterized in that: The details of vessel diameter measurement are as follows: After the centerline is extracted, the system interpolates the vessel contour and fits the vessel edge with a B-spline curve to ensure accurate reconstruction of the vessel morphology. To further improve the accuracy of diameter calculation, especially in areas with curved or narrow vessels, an improved MoM (Measure of Match) model is used. (1) Determination of blood vessel direction: In blood vessel contour fitting, the system first determines the direction of the blood vessel through curve fitting and Hessian matrix optimization. Assuming that point P is the current center point, the slope θ of the local fitting line segment at point P is weighted and combined. p1 The eigenvector θ corresponding to the smaller eigenvalue of the Hessian matrix p2 As the blood vessel direction θ p , defined as follows: i p =θ p1 +(1-α)θ p2 (2) Calculation of blood vessel diameter: The system calculates the vessel diameter based on the distance d between the left and right boundary points P1 and P2. The improved MoM model introduces an optimization of the matching degree of bilateral boundary points, making the boundary features on both sides clearer. The diameter calculation formula is: Blood vessel diameter = d l +d r Among them, d l and d r Respectively represent the distance from the center point P to the left and right boundary points P1 and P2, and the diameter of the blood vessel is obtained by summing them. (3) Edge matching measurement: The improved MoM algorithm maximizes edge matching by matching the actual edge of the blood vessel with the fitted curve. To improve the accuracy of the calculation, the algorithm calculates the matching degree on the left and right sides separately: Where m is the maximum search distance, f1(t+i) and f2(t+i) are the fitted left and right curves, g1(t+i) and g2(t+i) are the left and right vascular edge points, respectively. By calculating the deviation between the fitted curve and the actual edge, the improved MoM algorithm can maximize the matching degree of the left and right vascular edges and further optimize the accuracy of the vascular diameter. (4) Determination of blood vessel diameter: Finally, the system determines the optimal boundary point positions P1 and P2 on the left and right sides by maximizing the MoM values ​​(matching degree) on the left and right sides: The precise blood vessel diameter can be calculated from this.

7. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 6, characterized in that: In the stenosis rate calculation step, the reference diameter is calculated based on the average diameter of the adjacent healthy blood vessel regions before and after the stenosis region.

8. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 7, characterized in that: In the automatic stenosis correction step, B-spline curve interpolation is performed on the normal blood vessel reference contour in the stenosis area to smooth the corrected blood vessel contour.

9. The DSA coronary artery stenosis analysis method based on matching degree measurement according to claim 8, characterized in that: The method also includes a user interface function that supports manual adjustment of the stenosis area and dynamic display of the stenosis degree.

Images