Coronary blood vessel recognition method and device, electronic equipment and storage medium

Abstract

The embodiment of the invention discloses a coronary blood vessel recognition method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a coronary artery blood vessel image, and extracting a coronary artery center line from the coronary artery blood vessel image; based on the coronary artery center lines, segmenting coronary artery blood vessels in the coronary artery blood vessel image to obtain a plurality of blood vessel segments, and detecting a left coronary artery center line and a right coronary artery center line in the coronary artery center lines; based on the left coronary artery center line, multiple left coronary artery blood vessel segments corresponding to the left coronary artery center line in the multiple blood vessel segments are identified, and names corresponding to the multiple left coronary artery blood vessel segments are obtained; based on the right coronary artery center line, multiple right coronary artery blood vessel segments corresponding to the right coronary artery center line in the multiple blood vessel segments are recognized, and names corresponding to the multiple right coronary artery blood vessel segments are obtained. According to the technical scheme of the embodiment of the invention, the accuracy of coronary blood vessel recognition can be remarkably improved.

Description

Technical Field

[0001] Embodiments of the present invention relate to the field of image processing technologies, and in particular, to a coronary artery blood vessel recognition method, apparatus, electronic device, and storage medium. Background Art

[0002] Currently, computed tomography angiography (CTA) technology is one of the most commonly used imaging methods for diagnosing cardiovascular diseases. On this basis, in order to achieve an accurate evaluation of coronary artery blood vessels, it is crucial to identify each coronary artery blood vessel in a CTA image.

[0003] However, the existing recognition solutions for coronary artery blood vessels have the problem of low accuracy. Summary of the Invention

[0004] Embodiments of the present invention provide a coronary artery blood vessel recognition method, apparatus, electronic device, and storage medium to achieve accurate recognition of coronary artery blood vessels.

[0005] According to one aspect of the present invention, a coronary artery blood vessel recognition method is provided, which may include:

[0006] Obtain a coronary artery blood vessel image, and extract a coronary artery centerline from the coronary artery blood vessel image;

[0007] Based on the coronary artery centerline, segment the coronary artery blood vessels in the coronary artery blood vessel image to obtain multiple blood vessel segments, and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline;

[0008] Based on the left coronary artery centerline, identify multiple left coronary artery blood vessel segments corresponding to the left coronary artery centerline among the multiple blood vessel segments, and obtain the names corresponding to the multiple left coronary artery blood vessel segments respectively;

[0009] Based on the right coronary artery centerline, identify multiple right coronary artery blood vessel segments corresponding to the right coronary artery centerline among the multiple blood vessel segments, and obtain the names corresponding to the multiple right coronary artery blood vessel segments respectively.

[0010] According to another aspect of the present invention, a coronary artery blood vessel recognition apparatus is provided, which may include:

[0011] A coronary artery centerline extraction module, configured to obtain a coronary artery blood vessel image and extract a coronary artery centerline from the coronary artery blood vessel image;

[0012] A right coronary artery centerline detection module, configured to segment the coronary artery blood vessels in the coronary artery blood vessel image based on the coronary artery centerline to obtain multiple blood vessel segments, and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline;

[0013] The left coronary artery vessel segment name obtaining module is used to identify multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among multiple vessel segments based on the left coronary artery centerline, and obtain the names corresponding to the multiple left coronary artery vessel segments respectively;

[0014] The right coronary artery vessel segment name obtaining module is used to identify multiple right coronary artery vessel segments corresponding to the right coronary artery centerline among multiple vessel segments based on the right coronary artery centerline, and obtain the names corresponding to the multiple right coronary artery vessel segments respectively.

[0015] According to another aspect of the present invention, there is provided an electronic device, which may include:

[0016] At least one processor; and

[0017] A memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program executable by the at least one processor. When the computer program is executed by the at least one processor, the at least one processor is caused to implement the coronary artery vessel recognition method provided by any embodiment of the present invention.

[0019] According to another aspect of the present invention, there is provided a computer-readable storage medium, on which computer instructions are stored. When the computer instructions are used to cause a processor to execute, the coronary artery vessel recognition method provided by any embodiment of the present invention is implemented.

[0020] In the technical solution of the embodiment of the present invention, by acquiring a coronary artery vessel image and extracting the coronary artery centerline from the coronary artery vessel image; then, based on the coronary artery centerline, segmenting the coronary artery vessels in the coronary artery vessel image to obtain multiple vessel segments. The segmentation process of the coronary artery vessels helps to improve the accuracy of coronary artery vessel recognition, and to detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline. The distinction between the left and right coronary artery centerlines helps to achieve targeted recognition of the left coronary artery vessels and the right coronary artery vessels; then, based on the left coronary artery centerline, identifying multiple left coronary artery vessel segments among the multiple vessel segments, and obtaining the names corresponding to the multiple left coronary artery vessel segments respectively, thus completing the recognition process of the left coronary artery vessels, and based on the right coronary artery centerline, identifying multiple right coronary artery vessel segments among the multiple vessel segments, and obtaining the names corresponding to the multiple right coronary artery vessel segments respectively, thus completing the recognition process of the right coronary artery vessels. The above technical solution, through the segmentation process of the coronary artery vessels, and on this basis, combined with the targeted recognition of the left coronary artery vessel segments and the right coronary artery vessel segments in the multiple vessel segments, compared with recognizing the coronary artery vessels as a whole, the above more refined recognition process can significantly improve the accuracy of coronary artery vessel recognition.

[0021] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it used to limit the scope of the present invention. Other features of the present invention will become readily apparent from the following description. Description of the Drawings

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0023] Figure 1a is the first schematic diagram of the coronary artery vessel segmentation of SCCT provided according to an embodiment of the present invention;

[0024] Figure 1b is the second schematic diagram of the coronary artery vessel segmentation of SCCT provided according to an embodiment of the present invention;

[0025] Figure 2 is the flowchart of a coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0026] Figure 3 is the flowchart of another coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0027] Figure 4 is the flowchart of yet another coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0028] Figure 5 is the flowchart of still another coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0029] Figure 6 is the flowchart of still another coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0030] Figure 7 is the flowchart of an optional example in still another coronary artery vessel recognition method provided according to an embodiment of the present invention;

[0031] Figure 8 is the structural block diagram of a coronary artery vessel recognition device provided according to an embodiment of the present invention;

[0032] Figure 9 is the structural schematic diagram of an electronic device for implementing the coronary artery vessel recognition method of the embodiment of the present invention. Detailed Embodiments

[0033] To enable those skilled in the art to better understand the solution of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0034] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily need to describe a specific order or sequence. It should be understood that such used data can be interchanged under appropriate circumstances so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. The same is true for cases such as "target" and "original", which will not be elaborated here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0035] Before introducing the embodiments of the present invention, an exemplary description of the application scenarios of the embodiments of the present invention will be given first. Exemplarily, in combination with clinical practice experience, referring to the coronary artery vessel division / segmentation standard given by the Society of Cardiovascular Computed Tomography (SCCT), the coronary artery vessels can be divided into Figure 1a 18 categories as shown in and Table 1, which specifically may include the proximal right coronary artery (pRCA), the middle right coronary artery (mRCA), the distal right coronary artery (dRCA), the right coronary artery origin posterior descending branch (R-PDA), the left main trunk (LM), the proximal left anterior descending artery (pLAD), the middle left anterior descending artery (mLAD), the distal left anterior descending artery (dLAD), the first diagonal branch (D1), the second diagonal branch (D2), the proximal left circumflex artery (pLCX), the first obtuse marginal branch (OM1), the middle and distal left circumflex artery (LCx), the second obtuse marginal branch (OM2), the left circumflex artery origin posterior descending branch (L-PDA), the right coronary artery origin posterior lateral branch (R-PLB), the intermediate branch (RI), and the left circumflex artery origin posterior lateral branch (L-PLB).

[0036] On this basis, in order to more vividly understand various coronary artery vessels, further reference can be made to Figure 1b , which three-dimensionally shows pRCA, mRCA, LM, pLAD, mLAD, dLAD, pLCX, and L-PLB (i.e., Figure 1bin the PLB), and compared with Figure 1a , it also shows the distal left circumflex artery (dLCX) in the mid - distal segment of the left circumflex artery (LCx).

[0037] Table 1 SCCT coronary artery segmentation criteria

[0038]

[0039]

[0040] Next, the coronary artery recognition process will be elaborated in detail.

[0041] Figure 2 is a flowchart of a coronary artery recognition method provided by an embodiment of the present invention. This embodiment is applicable to the situation of coronary artery recognition, especially applicable to the situation of coronary artery recognition based on the SCCT coronary artery segmentation standard. This method can be executed by the coronary artery recognition device provided by the embodiment of the present invention. The device can be implemented in software and / or hardware, and can be integrated on an electronic device, which can be various user terminals or servers.

[0042] See Figure 2 , the method of the embodiment of the present invention specifically includes the following steps:

[0043] S110. Obtain a coronary artery image and extract the coronary artery centerline from the coronary artery image.

[0044] Among them, the coronary artery image can be understood as an image obtained by imaging the coronary artery using imaging techniques such as CTA. Obtain the coronary artery image.

[0045] Since the coronary artery image is obtained by imaging the coronary artery, the coronary artery image can reflect the coronary artery. Therefore, the coronary artery centerline of the coronary artery can be extracted from the coronary artery image. On this basis, combined with the application scenarios that the embodiment of the present invention may involve, the coronary arteries in the human body can be divided into left coronary arteries and right coronary arteries. Then, the coronary artery centerline extracted from the coronary artery image can include the left coronary artery centerline corresponding to the left coronary artery and the right coronary artery centerline corresponding to the right coronary artery.

[0046] S120. Based on the coronary artery centerline, segment the coronary artery in the coronary artery image to obtain multiple vascular segments, and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline.

[0047] Among them, based on the coronary artery centerline, segment the coronary artery in the coronary artery image to obtain multiple vascular segments. The segmentation process of the coronary artery effectively improves the accuracy of subsequent coronary artery recognition. Exemplarily, such as Figure 1aAs shown, each branch in the diagram is a blood vessel segment, and the white dashed line that cuts the main branch also indicates that the main branch is divided into individual blood vessel segments.

[0048] In practical applications, optionally, when segmenting the coronary blood vessels in a coronary blood vessel image, the coronary blood vessels can be first segmented from the coronary blood vessel image, and then, based on the coronary centerline, the segmented coronary blood vessels are segmented to obtain multiple blood vessel segments. The segmentation process of the coronary blood vessels helps to eliminate the background regions irrelevant to the coronary blood vessels in the coronary blood vessel image, thereby improving the accuracy of coronary blood vessel segmentation.

[0049] To improve the accuracy of coronary blood vessel recognition, in addition to segmenting the coronary blood vessels, the left coronary centerline and the right coronary centerline can also be detected from the coronary centerline. The distinction between the left and right coronary centerlines helps to achieve targeted recognition of the left coronary blood vessels and the right coronary blood vessels.

[0050] S130. Based on the left coronary centerline, identify multiple left coronary blood vessel segments corresponding to the left coronary centerline among the multiple blood vessel segments, and obtain the names corresponding to the multiple left coronary blood vessel segments respectively.

[0051] Among them, the left coronary blood vessel segment can be understood as the blood vessel segment obtained by segmenting the left coronary blood vessels. Determine multiple left coronary blood vessel segments from the multiple blood vessel segments. For example, the blood vessel segments corresponding to the left coronary centerline among the multiple blood vessel segments are used as the left coronary blood vessel segments, and thus multiple left coronary blood vessel segments are obtained.

[0052] On this basis, further, identify the multiple left coronary blood vessel segments based on the left coronary centerline respectively, and obtain the names corresponding to the multiple left coronary blood vessel segments respectively. So far, the recognition of the left coronary blood vessels is completed.

[0053] S140. Based on the right coronary centerline, identify multiple right coronary blood vessel segments corresponding to the right coronary centerline among the multiple blood vessel segments, and obtain the names corresponding to the multiple right coronary blood vessel segments respectively.

[0054] Among them, the right coronary blood vessel segment can be understood as the blood vessel segment obtained by segmenting the right coronary blood vessels. Determine multiple right coronary blood vessel segments from the multiple blood vessel segments. For example, the blood vessel segments corresponding to the right coronary centerline among the multiple blood vessel segments are used as the right coronary blood vessel segments, and thus multiple right coronary blood vessel segments are obtained.

[0055] On this basis, further, identify the multiple right coronary blood vessel segments based on the right coronary centerline respectively, and obtain the names corresponding to the multiple right coronary blood vessel segments respectively. So far, the recognition of the right coronary blood vessels is completed.

[0056] In the technical solution of the embodiment of the present invention, a coronary artery blood vessel image is obtained, and a coronary artery center line is extracted from the coronary artery blood vessel image; then, based on the coronary artery center line, the coronary artery blood vessels in the coronary artery blood vessel image are segmented to obtain multiple blood vessel segments. The segmentation of the coronary artery blood vessels helps to improve the accuracy of coronary artery blood vessel recognition and detect the left coronary artery center line and the right coronary artery center line in the coronary artery center line. The distinction between the left and right coronary artery center lines helps to achieve targeted recognition of the left coronary artery blood vessels and the right coronary artery blood vessels; then, based on the left coronary artery center line, multiple left coronary artery blood vessel segments in the multiple blood vessel segments are recognized, and the names corresponding to the multiple left coronary artery blood vessel segments are obtained, thus completing the recognition process of the left coronary artery blood vessels. And based on the right coronary artery center line, multiple right coronary artery blood vessel segments in the multiple blood vessel segments are recognized, and the names corresponding to the multiple right coronary artery blood vessel segments are obtained, thus completing the recognition process of the right coronary artery blood vessels. In the above technical solution, through the segmentation of the coronary artery blood vessels and the targeted recognition of the left coronary artery blood vessel segments and the right coronary artery blood vessel segments in the multiple blood vessel segments on this basis, compared with recognizing the coronary artery blood vessels as a whole, the above more refined recognition process can significantly improve the accuracy of coronary artery blood vessel recognition.

[0057] Figure 3 FIG. is a flowchart of another coronary artery blood vessel recognition method provided in the embodiment of the present invention. This embodiment is optimized based on the above technical solutions. In this embodiment, optionally, based on the coronary artery center line, the coronary artery blood vessels in the coronary artery blood vessel image are segmented to obtain multiple blood vessel segments, including: detecting multiple center line nodes on the coronary artery center line, where the center line nodes represent the connection points between the main branches and branches or between branches in the coronary artery blood vessels; based on the multiple center line nodes, the coronary artery blood vessels in the coronary artery blood vessel image are segmented to obtain multiple blood vessel segments. Among them, the explanations of the same or corresponding terms as those in the above embodiments are not repeated here.

[0058] See Figure 3 , the method of this embodiment may specifically include the following steps:

[0059] S210. Obtain a coronary artery blood vessel image, and extract a coronary artery center line from the coronary artery blood vessel image.

[0060] S220. Detect the left coronary artery center line and the right coronary artery center line in the coronary artery center line, and multiple center line nodes on the coronary artery center line, where the center line nodes represent the connection points between the main branches and branches or between branches in the coronary artery blood vessels.

[0061] Among them, the center line node can be understood as the pixel point on the coronary artery center line used to represent the connection point between the main branch and the branch or between branches in the coronary artery blood vessels, that is, the center line node is the bifurcation point on the coronary artery center line, and each bifurcation on the coronary artery center line corresponds to its own center line node.

[0062] From another perspective, each connection point between each branch connected to the main branch of the coronary artery and the main branch is a centerline node. Moreover, when there are further branches bifurcated from a certain branch, the connection point between the bifurcated branch and the branch is also a centerline node.

[0063] Exemplarily, Figure 1a the intersection points of 5, 11, 17 and 6, as well as the intersection points of 3, 4 and 16 in [[ ]] are all centerline nodes. Of course, Figure 1a there are many other centerline nodes in [[ ]], which will not be further exemplified herein.

[0064] Detect multiple centerline nodes from the coronary artery centerline.

[0065] In practical applications, optionally, the detection process of the centerline nodes can be implemented based on algorithms such as region growing algorithm and Ramer algorithm, which is not specifically limited herein. It should be noted that when using the Ramer algorithm to detect centerline nodes, an appropriate distance threshold needs to be selected. An overly large distance threshold is likely to cause information loss, while an overly small distance threshold is likely to cause excessive iteration times, resulting in too long calculation time.

[0066] S230. Segment the coronary artery in the coronary artery image based on multiple centerline nodes to obtain multiple vascular segments.

[0067] Among them, segment the coronary artery based on multiple centerline nodes to obtain multiple vascular segments. Combining with the application scenarios that the embodiments of the present invention may involve, exemplarily, division can be performed respectively at each centerline node on the coronary artery, so as to divide the coronary artery into multiple vascular segments.

[0068] S240. Identify multiple left coronary artery vascular segments corresponding to the left coronary artery centerline among the multiple vascular segments based on the left coronary artery centerline, and obtain the names corresponding to the multiple left coronary artery vascular segments respectively.

[0069] S250. Identify multiple right coronary artery vascular segments corresponding to the right coronary artery centerline among the multiple vascular segments based on the right coronary artery centerline, and obtain the names corresponding to the multiple right coronary artery vascular segments respectively.

[0070] The technical solution of the embodiments of the present invention detects multiple centerline nodes from the coronary artery centerline, and then performs segmentation processing on the coronary artery based on the multiple centerline nodes. The detection and application of the centerline nodes ensure the accuracy of the coronary artery segmentation.

[0071] An optional technical solution for detecting multiple centerline nodes on the coronary artery centerline includes:

[0072] For the current pixel point on the coronary artery centerline, dilation is performed on the coronary artery centerline based on this current pixel point to obtain dilated pixel points;

[0073] Obtain a preset pixel number threshold, and determine whether the current pixel point is a centerline node according to the numerical relationship between the number of dilated pixel points and the pixel number threshold;

[0074] When there is a next pixel point of the current pixel point on the coronary artery centerline, take the next pixel point as the current pixel point, and repeat the step of performing dilation on the coronary artery centerline based on the current pixel point to obtain dilated pixel points, so as to obtain multiple centerline nodes.

[0075] Among them, the current pixel point can be understood as the currently processed pixel point on the coronary artery centerline. Dilation is performed on the coronary artery centerline based on this current pixel point to obtain dilated pixel points.

[0076] The pixel number threshold can be understood as a preset threshold related to the number of pixel points, and in particular can be understood as a threshold for determining whether the current pixel point is a centerline node. Combining the application scenarios that the embodiments of the present invention may involve, optionally, since the dilated pixel points can be pixel points adjacent to the current pixel point, and the number of pixel points adjacent to the centerline node is usually at least 3, the pixel number threshold can be preset accordingly, for example, preset to 2 or 3, which is not specifically limited herein. Obtain the pixel number threshold.

[0077] Determine whether the current pixel point is a centerline node according to the numerical relationship between the number of dilated pixel points and the pixel number threshold. Exemplarily, continuing with the above example, when the pixel number threshold is 2, it can be determined that the current pixel point is a centerline node when the number of dilated pixel points > 2; when the pixel number threshold is 3, it can be determined that the current pixel point is a centerline node when the number of dilated pixel points ≥ 3; and so on, which is not specifically limited herein.

[0078] Update the next pixel point of the current pixel point to the current pixel point, and repeat the above steps, so that multiple centerline nodes can be detected from the coronary artery centerline.

[0079] The above technical solution realizes the accurate detection of the centerline nodes, which is the key to ensuring the accuracy of coronary artery segmentation.

[0080] Figure 4It is a flowchart of another coronary artery recognition method provided in an embodiment of the present invention. This embodiment is optimized based on the above technical solutions. In this embodiment, optionally, detecting the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline may include: determining whether there is a target right distance less than a preset right distance threshold in the right distances between the two coronary artery centerlines and the right coronary sinus respectively, and whether there is a target left distance less than a preset left distance threshold in the left distances between the two coronary artery centerlines and the left coronary sinus respectively; if so, taking the coronary artery centerline corresponding to the target right distance as the right coronary artery centerline, and taking the coronary artery centerline corresponding to the target left distance as the left coronary artery centerline; otherwise, segmenting the two coronary artery centerlines respectively, and taking the coronary artery centerline that is segmented more times as the left coronary artery centerline, and taking the coronary artery centerline that is segmented less times as the right coronary artery centerline. Among them, the explanations of the same or corresponding terms as those in the above embodiments will not be elaborated here.

[0081] See Figure 4 , the method of this embodiment may specifically include the following steps:

[0082] S310. Obtain a coronary artery image, extract the coronary artery centerline from the coronary artery image, and segment the coronary artery in the coronary artery image based on the coronary artery centerline to obtain a plurality of vascular segments.

[0083] S320. Determine whether there is a target right distance less than a preset right distance threshold in the right distances between the two coronary artery centerlines and the right coronary sinus respectively, and whether there is a target left distance less than a preset left distance threshold in the left distances between the two coronary artery centerlines and the left coronary sinus respectively.

[0084] Among them, in the case where both the left and right coronary arteries originate normally, the left coronary artery originates from the left coronary sinus and the right coronary artery originates from the right coronary sinus. In other words, the left coronary artery centerline is closer to the left coronary sinus and the right coronary artery centerline is closer to the right coronary sinus. Therefore, it is possible to determine whether the left and right coronary arteries originate abnormally according to the distances between the two coronary artery centerlines and the left and right coronary sinuses, and on this basis, distinguish the left coronary artery centerline from the right coronary artery centerline.

[0085] Specifically, the number of coronary artery centerlines extracted from the coronary artery image is two, one of which is the left coronary artery centerline and the other is the right coronary artery centerline.

[0086] The right distance threshold can be understood as a preset threshold related to the right separation distance, where the right separation distance can be understood as the separation distance between each of the two coronary artery centerlines and the right coronary sinus respectively. Calculate the right separation distances between the two coronary artery centerlines and the right coronary sinus respectively, and then determine whether there is a target right separation distance less than the right distance threshold among the two right separation distances. Considering the application scenarios that the embodiments of the present invention may involve, when the target right separation distance exists, the number of target right separation distances may be 1 or 2, which is related to the actual situation and is not specifically limited herein.

[0087] The left distance threshold can be understood as a preset threshold related to the left separation distance, where the left separation distance can be understood as the separation distance between each of the two coronary artery centerlines and the left coronary sinus respectively. Calculate the left separation distances between the two coronary artery centerlines and the left coronary sinus respectively, and then determine whether there is a target left separation distance less than the left distance threshold among the two left separation distances. Considering the application scenarios that the embodiments of the present invention may involve, when the target left separation distance exists, the number of target left separation distances may be 1 or 2, which is related to the actual situation and is not specifically limited herein.

[0088] S330. If so, use the coronary artery centerline corresponding to the target right separation distance as the right coronary artery centerline, and use the coronary artery centerline corresponding to the target left separation distance as the left coronary artery centerline.

[0089] Among them, when both the target right separation distance and the target left separation distance exist, referring to the above example, there is usually 1 target right separation distance and 1 target left separation distance, indicating that one of the two coronary artery centerlines is very close to the right coronary sinus and the other is very close to the left coronary sinus, that is, it indicates that both the left and right coronary arteries originate normally. Then, the coronary artery centerline corresponding to the target right separation distance can be used as the right coronary artery centerline, and the coronary artery centerline corresponding to the target left separation distance can be used as the left coronary artery centerline.

[0090] S340. Otherwise, divide the two coronary artery centerlines respectively, and use the coronary artery centerline that is divided more times among the two coronary artery centerlines as the left coronary artery centerline, and the coronary artery centerline that is divided less times as the right coronary artery centerline.

[0091] Among them, when at least one of the target right separation distance and the target left separation distance does not exist, referring to the above example, there may be 2 target right separation distances, 2 target left separation distances, 1 target right separation distance and 0 target left separation distances, or 0 target right separation distances and 1 target left separation distance, etc., which indicates that at least one of the left and right coronary arteries has an abnormal origin.

[0092] On this basis, considering that the left coronary artery is longer than the right coronary artery, that is, the center line of the left coronary artery is longer than that of the right coronary artery, the center lines of the two coronary arteries can be segmented respectively. Then, the coronary artery center line that is segmented more times among the two coronary artery center lines is used as the center line of the left coronary artery, and the coronary artery center line that is segmented less times is used as the center line of the right coronary artery. In practical applications, the coronary artery center line can be segmented based on the center line nodes.

[0093] S350. Based on the center line of the left coronary artery, identify multiple left coronary artery segments corresponding to the center line of the left coronary artery among multiple blood vessel segments, and obtain the names corresponding to the multiple left coronary artery segments respectively.

[0094] S360. Based on the center line of the right coronary artery, identify multiple right coronary artery segments corresponding to the center line of the right coronary artery among multiple blood vessel segments, and obtain the names corresponding to the multiple right coronary artery segments respectively.

[0095] The technical solution of the embodiment of the present invention determines whether the left and right coronary arteries have abnormal origins by calculating the distances between the center lines of the two coronary arteries and the left and right coronary sinuses respectively, and accurately divides the center lines of the left and right coronary arteries for the two cases of normal origin and abnormal origin respectively.

[0096] An optional technical solution, the above coronary artery identification method further includes:

[0097] Calculate the first distance between the center line of the left coronary artery and the right coronary sinus, and determine whether the left coronary artery represented by the center line of the left coronary artery originates from the right coronary sinus or the pulmonary artery according to the numerical relationship between the first distance and a preset first distance threshold; and / or

[0098] Calculate the second distance between the center line of the right coronary artery and the left coronary sinus, and determine whether the right coronary artery represented by the center line of the right coronary artery originates from the left coronary sinus or the pulmonary artery according to the numerical relationship between the second distance and a preset second distance threshold.

[0099] Among them, the first distance threshold can be understood as a preset threshold related to the first distance. In the case where the left coronary artery represented by the center line of the left coronary artery has an abnormal origin, calculate the first distance between the center line of the left coronary artery and the right coronary sinus, and then determine the origin situation of the left coronary artery according to the numerical relationship between the first distance and the first distance threshold. Exemplarily, when the first distance < the first distance threshold, it is determined that the left coronary artery originates from the right coronary sinus, otherwise it originates from the pulmonary artery.

[0100] Among them, the second distance threshold can be understood as a preset threshold related to the second distance. In the case of the abnormal origin of the right coronary artery characterized by the center line of the right coronary artery, the second distance between the center line of the right coronary artery and the left coronary sinus is calculated, and then the origin of the right coronary artery is determined according to the numerical relationship between the second distance and the second distance threshold. Exemplarily, when the second distance < the second distance threshold, it is determined that the right coronary artery originates from the left coronary sinus, otherwise it originates from the pulmonary artery.

[0101] The above technical solution realizes the accurate identification of the abnormal origin of the left and right coronary arteries.

[0102] Figure 5 It is a flowchart of another coronary artery recognition method provided in an embodiment of the present invention. This embodiment is optimized based on the above technical solutions. In this embodiment, optionally, the names respectively corresponding to multiple right coronary artery segments include at least one of the right coronary artery, the proximal right coronary artery, the middle right coronary artery, the posterior descending branch of the right coronary origin, and the posterior lateral branch of the right coronary origin. The explanations of the same or corresponding terms as those in the above embodiments are not repeated here.

[0103] See Figure 5 , the method of this embodiment may specifically include the following steps:

[0104] S410. Obtain a coronary artery image and extract the coronary artery center line from the coronary artery image.

[0105] S420. Based on the coronary artery center line, segment the coronary arteries in the coronary artery image to obtain multiple vascular segments, and detect the left coronary artery center line and the right coronary artery center line in the coronary artery center line.

[0106] S430. Based on the left coronary artery center line, identify multiple left coronary artery segments corresponding to the left coronary artery center line in the multiple vascular segments, and obtain the names respectively corresponding to the multiple left coronary artery segments.

[0107] S440. Based on the right coronary artery center line, identify multiple right coronary artery segments corresponding to the right coronary artery center line in the multiple vascular segments, and obtain the names respectively corresponding to the multiple right coronary artery segments, where the names respectively corresponding to the multiple right coronary artery segments include at least one of the right coronary artery, the proximal right coronary artery, the middle right coronary artery, the posterior descending branch of the right coronary origin, and the posterior lateral branch of the right coronary origin.

[0108] Among them, after respectively identifying multiple right coronary artery segments, the obtained names may include at least one of the right coronary artery (RCA), the proximal right coronary artery (pRCA), the middle right coronary artery (mRCA), the posterior descending branch of the right coronary origin (R-PDA), and the posterior lateral branch of the right coronary origin (R-PLB).

[0109] The technical solution of the embodiment of the present invention realizes the accurate identification of multiple right coronary artery segments.

[0110] An optional technical solution is to identify multiple right coronary artery segments corresponding to the right coronary artery centerline among multiple artery segments, and obtain the names corresponding to the multiple right coronary artery segments respectively, including:

[0111] Determine multiple right coronary artery segments corresponding to the right coronary artery centerline among multiple artery segments;

[0112] For the starting pixel point corresponding to the right coronary sinus on the right coronary artery centerline, determine the farthest pixel point on the right coronary artery centerline that is the farthest from the starting pixel point, and based on the starting pixel point and the farthest pixel point, extract the main branch centerline from the right coronary artery centerline;

[0113] Name at least one right coronary artery segment corresponding to the main branch centerline among the multiple right coronary artery segments as the right coronary artery.

[0114] Among them, the starting pixel point can be understood as the pixel point corresponding to the right coronary sinus on the right coronary artery centerline. The farthest pixel point can be understood as the pixel point on the right coronary artery centerline that is the farthest from the starting pixel point. The main branch centerline can be understood as the part on the right coronary artery centerline corresponding to the main branch of the right coronary artery. Determine the farthest pixel point, and based on the starting pixel point and the farthest pixel point, extract the main branch centerline from the right coronary artery centerline. Here, the right coronary sinus is used as prior information to find the main branch.

[0115] Combined with the application scenarios that the embodiment of the present invention may involve, optionally, the farthest pixel point can be determined through the following steps: on the right coronary artery centerline, expand layer by layer outward with the starting pixel point as the center, and end when reaching the termination pixel point on the right coronary artery centerline; then, during the expansion process, obtain the farthest pixel point that is the farthest from the starting pixel point. On this basis, optionally, the main branch centerline can be considered to be obtained during the search for the farthest pixel point. This technical solution uses the Dijkstra algorithm to realize the fast and accurate determination of the farthest pixel point.

[0116] Further, name one or more right coronary artery segments corresponding to the main branch centerline among the multiple right coronary artery segments as the right coronary artery (RCA). Exemplarily, Figure 1a 1, 2, and 3 in are all RCAs, where 1 is the proximal right coronary artery (pRCA), 2 is the middle right coronary artery (mRCA), and 3 is the distal right coronary artery (dRCA). The identification processes of these 3 RCAs will be elaborated in detail later.

[0117] The above technical solution realizes the accurate identification of RCA.

[0118] On this basis, optionally, at least one right coronary artery segment corresponding to the center line of the main branch among multiple right coronary artery segments is named the right coronary artery, including:

[0119] Determine the inflection point with the largest curvature and the inflection point with the second largest curvature among the main branch pixel points on the main branch center line, and take the inflection point closest to the starting pixel point among the two determined inflection points as the first inflection point, and the inflection point second closest to the starting pixel point as the second inflection point;

[0120] Name the right coronary artery segment with the starting pixel point and the first inflection point as endpoints among multiple right coronary artery segments as the proximal right coronary artery, and name the right coronary artery segment with the first inflection point and the second inflection point as endpoints as the middle right coronary artery.

[0121] Among them, determining the inflection point with the largest curvature and the inflection point with the second largest curvature among the main branch pixel points on the main branch center line is the key to further dividing the RCA into pRCA, mRCA, and dRCA. Combining with the application scenarios that the embodiments of the present invention may involve, exemplarily, the two inflection points can be determined through the following steps: Based on polynomial fitting, fit the main branch pixel points on the main branch center line into a function curve, and place the function curve in a pre-defined coordinate system; in the coordinate system, calculate the inflection point with the largest curvature and the inflection point with the second largest curvature on the function curve through the second derivative.

[0122] Further, take the inflection point closest to the starting pixel point among the two determined inflection points as the first inflection point (i.e., inflection point 1), and the inflection point second closest to the starting pixel point as the second inflection point (i.e., inflection point 2). Then, name the main branch between inflection point 1 and the starting pixel point (i.e., the right coronary artery segment located between inflection point 1 and the starting pixel point) as pRCA, name the main branch between inflection point 1 and inflection point 2 as mRCA, and name the main branch other than pRCA and mRCA as dRCA.

[0123] The above technical solution realizes the accurate identification of pRCA, mRCA, and dRCA.

[0124] On this basis, optionally, based on the right coronary artery center line, identify multiple right coronary artery segments corresponding to the right coronary artery center line among multiple blood vessel segments, and the obtained names corresponding to the multiple right coronary artery segments further include:

[0125] Taking the starting pixel point as a reference, find all target blood vessel segments among the multiple right coronary artery segments that are located after the second inflection point;

[0126] Name the target blood vessel segment closest to the right coronary sinus among all target blood vessel segments as the posterior descending branch originating from the right coronary sinus, and name the target blood vessel segment with the longest length as the posterolateral branch originating from the right coronary sinus.

[0127] Among them, with the starting pixel point as the starting reference, the target vascular segment can be understood as the right coronary artery vascular segment among multiple right coronary artery vascular segments that is located after inflection point 2. Combining with the application scenarios that the embodiments of the present invention may involve, it can be specifically understood as the branch on the right coronary artery that is located after inflection point 2. For each target vascular segment among all the target vascular segments, that is, for each branch among all the branches located after inflection point 2, calculate the length of this branch, and the distance between the end point of this branch, especially the end point of this branch and the right coronary sinus. Then, name the branch with the shortest distance to the right coronary sinus among all the branches as the right coronary artery origin posterior descending branch (R-PDA), and name the branch with the longest length as the right coronary artery origin posterior lateral branch (R-PLB).

[0128] The above technical solution realizes the accurate identification of R-PDA and R-PLB.

[0129] In order to better understand the naming process of each right coronary artery vascular segment as a whole, the following can be described by way of specific examples. Exemplarily, first, the Dijkstra algorithm can be used to find the farthest pixel point on the right coronary artery center line from the right coronary sinus, and record the main branch obtained during the process of finding the farthest pixel point as RCA. Then, based on polynomial fitting, fit the main branch pixel points corresponding to RCA into a function curve, and place this function curve in a coordinate system to find the two inflection points with the largest curvature on this function curve through the second derivative. Here, the inflection point closest to the starting pixel point (i.e., the right coronary sinus) is regarded as inflection point 1, and the inflection point with the second closest distance to the starting pixel point is regarded as inflection point 2. Regard the main branch between inflection point 1 and the starting pixel point as pRCA, and regard the main branch between inflection point 1 and inflection point 2 as mRCA. Then, find all the branches after inflection point 2, and calculate the length of each branch among all the branches and the distance between the end point of this branch and the right coronary sinus. Regard the branch with the shortest distance to the right coronary sinus as R-PDA, and regard the branch with the longest length as R-PLB. Thus, the identification of each right coronary artery vascular segment is completed.

[0130] The above example realizes the accurate identification of each right coronary artery vascular segment in combination with the prior knowledge of cardiac anatomy.

[0131] Figure 6 It is a flowchart of still another coronary artery vascular identification method provided in the embodiments of the present invention. This embodiment is optimized based on the above technical solutions. In this embodiment, optionally, the names corresponding to multiple left coronary artery vascular segments include at least one of left main trunk, anterior descending branch segment, proximal anterior descending branch, middle anterior descending branch, distal anterior descending branch, first diagonal branch, second diagonal branch, first obtuse marginal branch, second obtuse marginal branch, circumflex branch segment, proximal circumflex branch, mid-distal circumflex branch, circumflex artery origin posterior descending branch, and circumflex artery origin posterior lateral branch. Among them, the explanations of the same or corresponding terms as those in the above embodiments will not be repeated here.

[0132] See Figure 6 , the method of this embodiment may specifically include the following steps:

[0133] S510. Obtain coronary artery images and extract the coronary artery centerline from the coronary artery images.

[0134] S520. Based on the coronary artery centerline, segment the coronary arteries in the coronary artery images to obtain multiple vascular segments, and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline.

[0135] S530. Based on the left coronary artery centerline, identify multiple left coronary artery segments corresponding to the left coronary artery centerline among the multiple vascular segments, and obtain the names corresponding to the multiple left coronary artery segments respectively. Among them, the names corresponding to the multiple left coronary artery segments respectively include at least one of the left main trunk, anterior descending branch segment, proximal anterior descending branch, middle anterior descending branch, distal anterior descending branch, first diagonal branch, second diagonal branch, first obtuse marginal branch, second obtuse marginal branch, circumflex branch segment, proximal circumflex branch, mid-distal circumflex branch, circumflex origin posterior descending branch, and circumflex origin posterior lateral branch.

[0136] Among them, after respectively identifying the multiple left coronary artery segments, the obtained names may include at least one of the left main trunk (LM), anterior descending branch segment (LAD), proximal anterior descending branch (pLAD), middle anterior descending branch (mLAD), distal anterior descending branch (dLAD), first diagonal branch (D1), second diagonal branch (D2), first obtuse marginal branch (OM1), second obtuse marginal branch (OM2), circumflex branch segment (LCX), proximal circumflex branch (pLCX), mid-distal circumflex branch (LCx), circumflex origin posterior descending branch (L-PDA), and circumflex origin posterior lateral branch (L-PLB).

[0137] S540. Based on the right coronary artery centerline, identify multiple right coronary artery segments corresponding to the right coronary artery centerline among the multiple vascular segments, and obtain the names corresponding to the multiple right coronary artery segments respectively.

[0138] The above technical solution realizes the accurate identification of multiple left coronary artery segments.

[0139] An optional technical solution, based on the left coronary artery centerline, identify multiple left coronary artery segments corresponding to the left coronary artery centerline among the multiple vascular segments, and obtain the names corresponding to the multiple left coronary artery segments respectively, including:

[0140] Determine multiple left coronary artery segments corresponding to the left coronary artery centerline among the multiple vascular segments;

[0141] Determine the trifurcation node on the left coronary artery centerline, and name the left coronary artery segment with the left coronary sinus and the trifurcation node as endpoints among the multiple left coronary artery segments as the left main trunk.

[0142] Among them, the trifurcation node can be understood as the pixel point on the center line of the left coronary artery that represents the connection point between the main branch and the three branches. Exemplarily, for example, it can be Figure 1a the intersection points of 5, 11, 17, and 6 in

[0143] The above technical solution realizes the accurate identification of the LM.

[0144] On this basis, optionally, based on the center line of the left coronary artery, identifying multiple left coronary artery segments corresponding to the center line of the left coronary artery among multiple blood vessel segments, and obtaining the names corresponding to the multiple left coronary artery segments respectively further includes:

[0145] Determining the left coronary artery segment with the longest length and the left coronary artery segment with the second longest length among the multiple left coronary artery segments, and naming the left coronary artery segment closer to the left ventricle among the two determined left coronary artery segments as the left anterior descending artery segment, and naming the left coronary artery segment closer to the left atrium as the circumflex artery segment.

[0146] Among them, to determine the left coronary artery segment with the longest length and the left coronary artery segment with the second longest length among the multiple left coronary artery segments, exemplarily, the Dijkstra algorithm described above can be used to find these two left coronary artery segments. Then, naming the left coronary artery segment closer to the left ventricle among the two left coronary artery segments as the left anterior descending artery segment (LAD), and naming the left coronary artery segment closer to the left atrium as the circumflex artery segment (LCX).

[0147] The above technical solution realizes the accurate identification of the LAD and LCX.

[0148] On this basis, optionally, based on the center line of the left coronary artery, identifying multiple left coronary artery segments corresponding to the center line of the left coronary artery among multiple blood vessel segments, and obtaining the names corresponding to the multiple left coronary artery segments respectively, further includes:

[0149] Determining two left coronary artery segments that point to the left coronary artery segment named the left anterior descending artery segment among the multiple left coronary artery segments, and screening out the first diagonal branch close to the left main trunk and the second diagonal branch far from the left main trunk from the two determined left coronary artery segments;

[0150] Taking the first diagonal branch and the second diagonal branch as boundaries, dividing the left coronary artery segment named the left anterior descending artery segment into three segments, and naming them as the proximal left anterior descending artery segment, the middle left anterior descending artery segment, and the distal left anterior descending artery segment respectively based on the distance from the left main trunk.

[0151] Among them, two left coronary artery segments that point to the LAD, especially the middle part of the LAD, are determined from multiple left coronary artery segments. Then, the left coronary artery segment closer to the LM, especially the end of the LM, among the two left coronary artery segments is used as the first diagonal branch (D1), and the left coronary artery segment farther from the LM, especially the end of the LM, is used as the second diagonal branch (D2).

[0152] Taking D1 and D2 as the boundaries, the LAD is divided into the proximal left anterior descending artery (pLAD), the middle left anterior descending artery (mLAD), and the distal left anterior descending artery (dLAD). Among them, the pLAD is close to the LM, and then followed by the mLAD and the dLAD in sequence.

[0153] The above technical solution realizes the accurate identification of D1 and D2, as well as pLAD, mLAD, and dLAD.

[0154] Another alternative is to identify multiple left coronary artery segments corresponding to the left coronary artery centerline among multiple artery segments, and obtain the names corresponding to the multiple left coronary artery segments, which further includes:

[0155] Determine two left coronary artery segments that point to the left coronary artery segment named the circumflex artery segment among multiple left coronary artery segments, and screen out the first obtuse marginal branch close to the left main trunk and the second obtuse marginal branch far from the left main trunk from the two determined left coronary artery segments;

[0156] Taking the first obtuse marginal branch and the second obtuse marginal branch as the boundaries, the left coronary artery segment named the circumflex artery segment is divided into two segments, and they are respectively named the proximal circumflex artery and the middle and distal circumflex artery based on the distance from the left main trunk.

[0157] Among them, two left coronary artery segments that point to the LCX, especially the middle part of the LCX, are determined from multiple left coronary artery segments. Then, the left coronary artery segment close to the LM among the two left coronary artery segments is named the first obtuse marginal branch (OM1), and the left coronary artery segment far from the LM is named the second obtuse marginal branch (OM2). Further, taking OM1 and OM2 as the boundaries, the LCX is divided into the proximal circumflex artery (pLCX) and the middle and distal circumflex artery (LCx). Among them, the pLCX is closer to the LM, and then comes the LCx.

[0158] The above technical solution realizes the accurate identification of OM1 and OM2, as well as pLCX and LCx.

[0159] On this basis, optionally, based on the left coronary artery centerline, identifying multiple left coronary artery segments corresponding to the left coronary artery centerline among multiple artery segments, and obtaining the names corresponding to the multiple left coronary artery segments further includes:

[0160] Taking the left coronary cusp as the starting pixel point, determine all target vessel segments after the left coronary artery vessel segment located in the mid-distal segment of the circumflex artery among multiple left coronary artery vessel segments;

[0161] Name the target vessel segment closest to the left coronary cusp among all target vessel segments as the posterior descending branch after the origin of the circumflex artery, and name the target vessel segment with the longest length as the posterior lateral branch after the origin of the circumflex artery.

[0162] Among them, taking the left coronary cusp as the starting reference, the target vessel segment can be understood as the left coronary artery vessel segment after the LCx among multiple left coronary artery vessel segments. Combining with the application scenarios that the embodiments of the present invention may involve, it can also be understood as a branch in the LCX. For each branch among all branches in all LCXs, calculate the length of the branch and the distance between the end point of the branch, especially the end point of the branch, and the left coronary cusp. Then name the branch closest to the left coronary cusp among all branches as the posterior descending branch after the origin of the circumflex artery (L-PDA), and name the branch with the longest length as the posterior lateral branch after the origin of the circumflex artery (L-PLB).

[0163] The above technical solution realizes the accurate identification of L-PDA and L-PLB.

[0164] To better understand the naming process of each left coronary artery vessel segment as a whole, the following can be exemplarily described with specific examples. Exemplarily, first, find the left coronary cusp and the trifurcation node, and regard the part from the trifurcation node to the left coronary cusp as the LM; find the left coronary artery vessel segments with the longest and the second longest lengths through the Dijkstra algorithm, and regard the one closer to the left ventricle among the two left coronary artery vessel segments as the LAD, and the one closer to the left atrium as the LCX. Then, in the LAD, find the ones pointing to the middle part between the LAD and the LCX as D1 and D2, where the one closer to the end of the LM is regarded as D1, and the one farther from the end of the LM is regarded as D2. Taking D1 and D2 as the boundaries, divide the LAD into pLAD, mLAD and dLAD, where the one closer to the LM is pLAD, and then mLAD and dLAD in turn. In the LCX, find the ones pointing to the middle part between the LAD and the LCX as OM1 and OM2, where the one closer to the LM is regarded as OM1, and the second is OM2. Taking OM1 and OM2 as the boundaries, divide the LCX into pLCX and LCx. Then, calculate the length of each branch in the LCX and the distance between the end point of the branch and the left coronary cusp, and regard the branch closest to the left coronary cusp as the L-PDA, and the branch with the longest length as the L-PLB. So far, the identification of each left coronary artery vessel segment is completed.

[0165] The above example realizes the accurate identification of each left coronary artery vessel segment in combination with the prior knowledge of cardiac anatomy.

[0166] To better understand the above technical solutions as a whole, the following provides an exemplary description in combination with specific examples. Exemplarily, refer to Figure 7 :

[0167] Step 1: Obtain a CTA image, that is, 3D data of the heart coronary arteries, and then define the coordinate axes of the 3D data of the heart coronary arteries to define the directions of the three-dimensional coordinates of the 3D data of the heart coronary arteries, which can be used for subsequent distance calculation and function curve fitting. Among them, the definitions of the coordinate axes of the coordinate system are as follows:

[0168] X-axis: The left-right direction of the human body is the X-axis, and the left half (i.e., the direction where the left hand is located) is the positive direction of the X-axis;

[0169] Y-axis: The direction from the chest to the back of the human body is the positive direction of the Y-axis;

[0170] Z-axis: The direction from the head to the feet of the human body is the positive direction of the Z-axis.

[0171] Step 2: Based on the deep learning model, implement heart and coronary artery detection (i.e., segmentation):

[0172] Use the pre-trained deep learning model to detect coronary arteries, as well as the left ventricle, left atrium, left coronary sinus, and right coronary sinus from the 3D data of the heart coronary arteries. Among them, the detection results of the coronary arteries are used to identify different types of coronary arteries, and the detection results of the left ventricle, left atrium, left coronary sinus, and right coronary sinus are used as prior information to assist in identifying different types of coronary arteries.

[0173] Step 3: Extract two coronary centerlines from the 3D data of the heart coronary arteries, and then based on the two coronary centerlines, segment the coronary arteries and determine whether the left and right coronary arteries have abnormal origins.

[0174] Step 4: According to the origin judgment results of the left and right coronary arteries, the two coronary centerlines can be divided into the left coronary centerline and the right coronary centerline.

[0175] Step 5: Based on the left coronary centerline and the right coronary centerline, and in combination with the SCCT coronary artery segmentation standard, name each vascular segment.

[0176] In the above example, based on the coronary centerline and combined with the prior knowledge of cardiac anatomy, it is possible to accurately identify different types of coronary arteries and accurately identify them in the case of abnormal coronary artery origins.

[0177] Figure 8The following is a structural block diagram of the coronary artery blood vessel recognition device provided in the embodiments of the present invention. This device is used to execute the coronary artery blood vessel recognition method provided in any of the above embodiments. This device and the coronary artery blood vessel recognition methods in the above embodiments belong to the same inventive concept. For the details not described in detail in the embodiments of the coronary artery blood vessel recognition device, reference can be made to the embodiments of the above coronary artery blood vessel recognition method. Refer to Figure 8 , specifically, this device may include: a coronary artery centerline extraction module 610, a right coronary artery centerline detection module 620, a left coronary artery blood vessel segment name obtaining module 630, and a right coronary artery blood vessel segment name obtaining module 640.

[0178] Among them, the coronary artery centerline extraction module 610 is used to obtain a coronary artery blood vessel image and extract the coronary artery centerline from the coronary artery blood vessel image;

[0179] The right coronary artery centerline detection module 620 is used to segment the coronary artery blood vessels in the coronary artery blood vessel image based on the coronary artery centerline to obtain a plurality of blood vessel segments, and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline;

[0180] The left coronary artery blood vessel segment name obtaining module 630 is used to identify a plurality of left coronary artery blood vessel segments corresponding to the left coronary artery centerline among the plurality of blood vessel segments based on the left coronary artery centerline, and obtain the names corresponding to the plurality of left coronary artery blood vessel segments respectively;

[0181] The right coronary artery blood vessel segment name obtaining module 640 is used to identify a plurality of right coronary artery blood vessel segments corresponding to the right coronary artery centerline among the plurality of blood vessel segments based on the right coronary artery centerline, and obtain the names corresponding to the plurality of right coronary artery blood vessel segments respectively.

[0182] Optionally, the right coronary artery centerline detection module 620 may include:

[0183] A centerline node detection sub-module, which is used to detect a plurality of centerline nodes on the coronary artery centerline, where the centerline nodes represent the connection points between the main branches and branches or between branches in the coronary artery blood vessels;

[0184] A blood vessel segment obtaining sub-module, which is used to segment the coronary artery blood vessels in the coronary artery blood vessel image based on the plurality of centerline nodes to obtain a plurality of blood vessel segments.

[0185] On this basis, optionally, the centerline node detection sub-module may include:

[0186] An expanded pixel point obtaining unit, which is used to expand the current pixel point on the coronary artery centerline on the coronary artery centerline to obtain an expanded pixel point;

[0187] A centerline node determination unit, configured to obtain a preset threshold of the number of pixel points, and determine whether a current pixel point is a centerline node according to the numerical relationship between the number of dilated pixel points and the threshold of the number of pixel points;

[0188] A centerline node obtaining unit, configured to, when there is a next pixel point of the current pixel point on the coronary artery centerline, use the next pixel point as the current pixel point, and repeatedly execute the step of dilating based on the current pixel point on the coronary artery centerline to obtain dilated pixel points, so as to obtain a plurality of centerline nodes.

[0189] Optionally, the right coronary artery centerline detection module 620 may include:

[0190] A target right distance determination sub-module, configured to determine whether there is a target right distance less than a preset right distance threshold in the right distances between two coronary artery centerlines and the right coronary sinus respectively, and whether there is a target left distance less than a preset left distance threshold in the left distances between the two coronary artery centerlines and the left coronary sinus respectively;

[0191] A left coronary artery centerline obtaining sub-module, configured to, if so, use the coronary artery centerline corresponding to the target right distance as the right coronary artery centerline, and use the coronary artery centerline corresponding to the target left distance as the left coronary artery centerline;

[0192] A right coronary artery centerline obtaining sub-module, configured to, otherwise, segment the two coronary artery centerlines respectively, and use the coronary artery centerline that is segmented more times among the two coronary artery centerlines as the left coronary artery centerline, and use the coronary artery centerline that is segmented less times as the right coronary artery centerline.

[0193] On this basis, optionally, the above-mentioned coronary artery vessel recognition device further includes:

[0194] A left coronary artery origin determination module, configured to calculate a first distance between the left coronary artery centerline and the right coronary sinus, and determine whether the left coronary artery represented by the left coronary artery centerline originates from the right coronary sinus or the pulmonary artery according to the numerical relationship between the first distance and a preset first distance threshold; and / or,

[0195] A right coronary artery origin determination module, configured to calculate a second distance between the right coronary artery centerline and the left coronary sinus, and determine whether the right coronary artery represented by the right coronary artery centerline originates from the left coronary sinus or the pulmonary artery according to the numerical relationship between the second distance and a preset second distance threshold.

[0196] Optionally, the names corresponding to multiple right coronary artery segments include at least one of the right coronary artery, the proximal right coronary artery, the middle right coronary artery, the posterior descending branch of the right coronary origin, and the posterior lateral branch of the right coronary origin.

[0197] On this basis, optionally, the right coronary artery segment name obtaining module 640 may include:

[0198] A right coronary artery segment determination sub-module, configured to determine multiple right coronary artery segments corresponding to the right coronary artery centerline among multiple artery segments;

[0199] A main branch centerline extraction sub-module, configured to, for the starting pixel point corresponding to the right coronary sinus on the right coronary artery centerline, determine the farthest pixel point on the right coronary artery centerline that is farthest from the starting pixel point, and based on the starting pixel point and the farthest pixel point, extract the main branch centerline from the right coronary artery centerline;

[0200] A right coronary artery naming sub-module, configured to name at least one right coronary artery segment corresponding to the main branch centerline among the multiple right coronary artery segments as the right coronary artery.

[0201] On this basis, optionally, the main branch centerline extraction sub-module may include:

[0202] A pixel point expansion unit, configured to expand layer by layer outward with the starting pixel point as the center on the right coronary artery centerline, and end when expanding to the ending pixel point on the right coronary artery centerline;

[0203] A farthest pixel point obtaining unit, configured to obtain the farthest pixel point that is farthest from the starting pixel point during the expansion process.

[0204] Another optionally, the right coronary artery naming sub-module may include:

[0205] A second inflection point obtaining unit, configured to determine the inflection point with the largest curvature and the inflection point with the second largest curvature among the main branch pixel points on the main branch centerline, and use the inflection point closest to the starting pixel point among the two determined inflection points as the first inflection point, and the inflection point second closest to the starting pixel point as the second inflection point;

[0206] A right coronary artery middle segment naming unit, configured to name the right coronary artery segment with the starting pixel point and the first inflection point as endpoints among the multiple right coronary artery segments as the proximal right coronary artery, and name the right coronary artery segment with the first inflection point and the second inflection point as endpoints as the middle right coronary artery segment.

[0207] On this basis, optionally, the second inflection point obtaining unit may include:

[0208] A function curve placement sub-unit, configured to fit the main branch pixel points on the main branch centerline into a function curve based on polynomial fitting, and place the function curve in a pre-defined coordinate system;

[0209] An inflection point calculation sub-unit, configured to calculate the inflection point with the largest curvature and the inflection point with the second largest curvature on the function curve in the coordinate system through the second derivative.

[0210] Optionally, the right coronary artery vessel segment name obtaining module 640 may further include:

[0211] A target vessel segment finding sub-module, configured to find all target vessel segments of the plurality of right coronary artery vessel segments located after the second inflection point based on the starting pixel point;

[0212] A right coronary origin posterior branch naming sub-module, configured to name the target vessel segment closest to the right coronary sinus among all the target vessel segments as the posterior descending branch of the right coronary origin, and name the target vessel segment with the longest length as the posterior branch of the right coronary origin.

[0213] Optionally, the names respectively corresponding to the plurality of left coronary artery vessel segments include at least one of the left main trunk, the anterior descending branch segment, the proximal anterior descending branch, the middle anterior descending branch, the distal anterior descending branch, the first diagonal branch, the second diagonal branch, the first obtuse marginal branch, the second obtuse marginal branch, the circumflex branch segment, the proximal circumflex branch, the middle and distal circumflex branch, the posterior descending branch of the circumflex origin, and the posterior branch of the circumflex origin.

[0214] On this basis, optionally, the left coronary artery vessel segment name obtaining module 630 may include:

[0215] A left coronary artery vessel segment determining unit, configured to determine a plurality of left coronary artery vessel segments corresponding to the left coronary artery center line among the plurality of vessel segments;

[0216] A left main trunk naming unit, configured to determine a trifurcation node on the left coronary artery center line, and name the left coronary artery vessel segment with the left coronary sinus and the trifurcation node as endpoints among the plurality of left coronary artery vessel segments as the left main trunk.

[0217] On this basis, optionally, the left coronary artery vessel segment name obtaining module 630 may further include:

[0218] A circumflex branch segment naming unit, configured to determine the left coronary artery vessel segment with the longest length and the left coronary artery vessel segment with the second longest length among the plurality of left coronary artery vessel segments, and name the left coronary artery vessel segment closer to the left ventricle among the two determined left coronary artery vessel segments as the anterior descending branch segment, and name the left coronary artery vessel segment closer to the left atrium as the circumflex branch segment.

[0219] On this basis, optionally, the left coronary artery vessel segment name obtaining module 630 may further include:

[0220] A second diagonal branch naming unit, configured to determine two left coronary artery vessel segments pointing to the left coronary artery vessel segment named the anterior descending branch segment among the plurality of left coronary artery vessel segments, and screen out the first diagonal branch close to the left main trunk and the second diagonal branch far from the left main trunk from the two determined left coronary artery vessel segments;

[0221] The distal segment naming unit of the left anterior descending artery is used to divide the left coronary artery segment named the left anterior descending artery segment into three segments with the first diagonal branch and the second diagonal branch as boundaries, and name them the proximal segment of the left anterior descending artery, the middle segment of the left anterior descending artery, and the distal segment of the left anterior descending artery respectively based on the distance from the left main trunk.

[0222] Optionally, the left coronary artery segment name obtaining module 630 may further include:

[0223] The second obtuse marginal branch naming unit is used to determine two left coronary artery segments among multiple left coronary artery segments that point to the left coronary artery segment named the circumflex artery segment, and screen out the first obtuse marginal branch close to the left main trunk and the second obtuse marginal branch far from the left main trunk from the two determined left coronary artery segments;

[0224] The middle and distal segment naming unit of the circumflex artery is used to divide the left coronary artery segment named the circumflex artery segment into two segments with the first obtuse marginal branch and the second obtuse marginal branch as boundaries, and name them the proximal segment of the circumflex artery and the middle and distal segments of the circumflex artery respectively based on the distance from the left main trunk.

[0225] On this basis, optionally, the left coronary artery segment name obtaining module 630 may further include:

[0226] The target vessel segment determining unit is used to determine all target vessel segments located after the left coronary artery segment named the middle and distal segments of the circumflex artery from multiple left coronary artery segments with the left coronary sinus as the starting pixel point;

[0227] The circumflex artery origin posterior branch naming unit is used to name the target vessel segment closest to the left coronary sinus among all target vessel segments as the posterior descending branch of the circumflex artery origin, and the target vessel segment with the longest length as the circumflex artery origin posterior branch.

[0228] Optionally, the right coronary artery centerline detection module 620 may include:

[0229] The coronary artery segmentation unit is used to segment the coronary artery from the coronary artery image;

[0230] The vessel segment obtaining unit is used to segment the coronary artery based on the coronary artery centerline to obtain multiple vessel segments.

[0231] The coronary artery vessel recognition device provided by the embodiment of the present invention obtains a coronary artery vessel image through a coronary artery centerline extraction module, and extracts the coronary artery centerline from the coronary artery vessel image; then, through a right coronary artery centerline detection module, based on the coronary artery centerline, segment processing of the coronary artery vessels in the coronary artery vessel image is performed to obtain multiple vessel segments. The segment processing of the coronary artery vessels helps to improve the accuracy of coronary artery vessel recognition and detect the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline. The distinction between the left and right coronary artery centerlines helps to achieve targeted recognition of the left coronary artery vessels and the right coronary artery vessels; then, through a left coronary artery vessel segment name obtaining module, based on the left coronary artery centerline, multiple left coronary artery vessel segments in the multiple vessel segments are recognized, and the names corresponding to the multiple left coronary artery vessel segments are obtained, thus completing the recognition process of the left coronary artery vessels; and, through a right coronary artery vessel segment name obtaining module, based on the right coronary artery centerline, multiple right coronary artery vessel segments in the multiple vessel segments are recognized, and the names corresponding to the multiple right coronary artery vessel segments are obtained, thus completing the recognition process of the right coronary artery vessels. The above device, through the segment processing of the coronary artery vessels, and on this basis, combined with the targeted recognition of the left coronary artery vessel segments and the right coronary artery vessel segments in the multiple vessel segments, compared with recognizing the coronary artery vessels as a whole, the above more refined recognition process can significantly improve the accuracy of coronary artery vessel recognition.

[0232] The coronary artery vessel recognition device provided by the embodiment of the present invention can execute the coronary artery vessel recognition method provided by any embodiment of the present invention, and has the corresponding function modules and beneficial effects for executing the method.

[0233] It should be noted that in the embodiments of the above coronary artery vessel recognition device, the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be realized; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0234] Figure 9 FIG. shows a schematic structural diagram of an electronic device 10 that can be used to implement the embodiments of the present invention. The electronic device is intended to represent various forms of digital computers, such as, laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as, personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are only examples and are not intended to limit the implementation of the present invention described herein and / or claimed.

[0235] As Figure 9As shown, the electronic device 10 includes at least one processor 11 and a memory communicatively connected to the at least one processor 11, such as read-only memory (ROM) 12, random access memory (RAM) 13, etc. Among them, the memory stores a computer program executable by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, ROM 12, and RAM 13 are connected to each other through a bus 14. The input / output (I / O) interface 15 is also connected to the bus 14.

[0236] Multiple components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disc, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0237] The processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the processor 11 include but are not limited to a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the coronary blood vessel recognition method.

[0238] In some embodiments, the coronary blood vessel recognition method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the coronary blood vessel recognition method described above can be executed. Alternatively, in other embodiments, the processor 11 can be configured to execute the coronary blood vessel recognition method by any other appropriate means (e.g., by means of firmware).

[0239] The various embodiments of the systems and techniques described above in this specification can be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0240] The computer programs for implementing the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general purpose computer, a special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The computer programs can be executed entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0241] In the context of the present invention, a computer-readable storage medium can be a tangible medium that can contain or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer-readable storage medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, the computer-readable storage medium can be a machine-readable signal medium. More specific examples of the machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0242] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device for displaying information to the user (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor); and a keyboard and a pointing device (e.g., a mouse or a trackball) through which the user can provide input to the electronic device. Other kinds of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0243] The systems and techniques described herein can be implemented in a computing system including backend components (e.g., as a data server), or a computing system including middleware components (e.g., an application server), or a computing system including frontend components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system including any combination of such backend components, middleware components, or frontend components. The components of the system can be interconnected by digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include: local area network (LAN), wide area network (WAN), blockchain network, and the Internet.

[0244] The computing system can include a client and a server. The client and the server are generally far from each other and typically interact through a communication network. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or a cloud host, which is a host product in the cloud computing service system, solving the defects of difficult management and weak business scalability existing in traditional physical hosts and VPS services.

[0245] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps recited in the present invention can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved, and no limitation is imposed herein.

[0246] The above specific embodiments do not constitute a limitation on the protection scope of the present invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A coronary artery blood vessel recognition method, characterized in that, Including: Obtaining a coronary artery blood vessel image, and extracting a coronary artery centerline from the coronary artery blood vessel image; Based on the coronary artery centerline, segmenting the coronary artery blood vessels in the coronary artery blood vessel image to obtain a plurality of blood vessel segments, and detecting a left coronary artery centerline and a right coronary artery centerline in the coronary artery centerline; Based on the left coronary artery centerline, identifying a plurality of left coronary artery blood vessel segments corresponding to the left coronary artery centerline among the plurality of blood vessel segments, and obtaining the names corresponding to the plurality of left coronary artery blood vessel segments respectively; Based on the right coronary artery centerline, identifying a plurality of right coronary artery blood vessel segments corresponding to the right coronary artery centerline among the plurality of blood vessel segments, and obtaining the names corresponding to the plurality of right coronary artery blood vessel segments respectively.

2. The method according to claim 1, wherein The segmenting the coronary artery blood vessels in the coronary artery blood vessel image based on the coronary artery centerline to obtain a plurality of blood vessel segments includes: Detecting a plurality of centerline nodes on the coronary artery centerline, where the centerline nodes represent connection points between main branches and branches or between branches in the coronary artery blood vessels; Based on the plurality of centerline nodes, segmenting the coronary artery blood vessels in the coronary artery blood vessel image to obtain a plurality of blood vessel segments.

3. The method according to claim 2, wherein The detecting the plurality of centerline nodes on the coronary artery centerline includes: For a current pixel point on the coronary artery centerline, performing dilation on the coronary artery centerline based on the current pixel point to obtain a dilated pixel point; Obtaining a preset pixel point quantity threshold, and determining whether the current pixel point is a centerline node according to the numerical relationship between the quantity of the dilated pixel points and the pixel point quantity threshold; In the case where there is a next pixel point of the current pixel point on the coronary artery centerline, taking the next pixel point as the current pixel point, and repeatedly executing the step of performing dilation on the coronary artery centerline based on the current pixel point to obtain a plurality of centerline nodes.

4. The method according to claim 1, characterized in that, The detecting the left coronary artery centerline and the right coronary artery centerline in the coronary artery centerline includes: Judging whether there is a target right distance less than a preset right distance threshold among the right distances between the two coronary artery centerlines and the right coronary sinus respectively, and whether there is a target left distance less than a preset left distance threshold among the left distances between the two coronary artery centerlines and the left coronary sinus respectively; If so, taking the coronary artery centerline corresponding to the target right distance as the right coronary artery centerline, and taking the coronary artery centerline corresponding to the target left distance as the left coronary artery centerline; Otherwise, respectively segmenting the two coronary artery centerlines, and taking the coronary artery centerline that is segmented more times among the two coronary artery centerlines as the left coronary artery centerline, and taking the coronary artery centerline that is segmented less times as the right coronary artery centerline.

5. The method according to claim 4, characterized in that It further includes: Calculating a first distance between the left coronary artery centerline and the right coronary sinus, and determining whether the left coronary artery represented by the left coronary artery centerline originates from the right coronary sinus or the pulmonary artery according to the numerical relationship between the first distance and a preset first distance threshold; and / or, Calculate a second distance between the centerline of the right coronary artery and the left coronary sinus, and determine whether the right coronary artery represented by the centerline of the right coronary artery originates from the left coronary sinus or the pulmonary artery according to the numerical relationship between the second distance and a preset second distance threshold.

6. The method according to claim 1, characterized in that, The names corresponding to the multiple right coronary artery segments respectively include at least one of the right coronary artery, the proximal right coronary artery, the middle right coronary artery, the posterior descending branch originating from the right coronary artery, and the posterior lateral branch originating from the right coronary artery.

7. The method according to claim 6, wherein Based on the centerline of the right coronary artery, identify multiple right coronary artery segments corresponding to the centerline of the right coronary artery among the multiple blood vessel segments, and obtain the names corresponding to the multiple right coronary artery segments respectively, including: Determine multiple right coronary artery segments corresponding to the centerline of the right coronary artery among the multiple blood vessel segments; For the starting pixel point corresponding to the right coronary sinus on the centerline of the right coronary artery, determine the farthest pixel point on the centerline of the right coronary artery that is farthest from the starting pixel point, and extract the main branch centerline from the centerline of the right coronary artery based on the starting pixel point and the farthest pixel point. Name at least one right coronary artery segment corresponding to the main branch centerline among the multiple right coronary artery segments as the right coronary artery.

8. The method according to claim 7, wherein The determination of the farthest pixel point on the centerline of the right coronary artery that is farthest from the starting pixel point includes: On the centerline of the right coronary artery, expand layer by layer outward with the starting pixel point as the center, and end when expanding to the ending pixel point on the centerline of the right coronary artery; During the expansion process, obtain the farthest pixel point that is farthest from the starting pixel point.

9. The method according to claim 7, wherein The naming of at least one right coronary artery segment corresponding to the main branch centerline among the multiple right coronary artery segments as the right coronary artery includes: Determine the inflection point with the largest curvature and the inflection point with the second largest curvature among the main branch pixel points on the main branch centerline, and use the inflection point closest to the starting pixel point among the two determined inflection points as the first inflection point, and the inflection point that is the second closest to the starting pixel point as the second inflection point; Name the right coronary artery segment with the starting pixel point and the first inflection point as endpoints among the multiple right coronary artery segments as the proximal right coronary artery, and name the right coronary artery segment with the first inflection point and the second inflection point as endpoints as the middle right coronary artery.

10. The method according to claim 9, characterized in that, The determination of the inflection point with the largest curvature and the inflection point with the second largest curvature among the main branch pixel points on the main branch centerline includes: Based on polynomial fitting, fit the main branch pixel points on the main branch centerline into a function curve, and place the function curve in a pre-defined coordinate system; In the coordinate system, calculate the inflection point with the largest curvature and the inflection point with the second largest curvature on the function curve through the second derivative.

11. The method according to claim 9, wherein Based on the centerline of the right coronary artery, identifying multiple right coronary artery segments corresponding to the centerline of the right coronary artery among the multiple blood vessel segments and obtaining the names corresponding to the multiple right coronary artery segments respectively further includes: Taking the starting pixel point as a reference, find all target blood vessel segments among the multiple right coronary artery segments that are located after the second inflection point; Name the target vessel segment closest to the right coronary sinus among all the target vessel segments as the posterior descending branch originating from the right coronary artery, and name the target vessel segment with the longest length as the posterolateral branch originating from the right coronary artery.

12. The method according to claim 1, wherein The respective names corresponding to the multiple left coronary artery vessel segments include at least one of the left main trunk, the anterior descending branch segment, the proximal anterior descending branch, the middle anterior descending branch, the distal anterior descending branch, the first diagonal branch, the second diagonal branch, the first obtuse marginal branch, the second obtuse marginal branch, the circumflex branch segment, the proximal circumflex branch, the mid-distal circumflex branch, the posterior descending branch originating from the circumflex branch, and the posterolateral branch originating from the circumflex branch.

13. The method according to claim 12, wherein Based on the left coronary artery centerline, identify the multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among the multiple vessel segments, and obtain the respective names corresponding to the multiple left coronary artery vessel segments, including: Determine the multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among the multiple vessel segments; Determine the trifurcation node on the left coronary artery centerline, and name the left coronary artery vessel segment with the left coronary sinus and the trifurcation node as endpoints among the multiple left coronary artery vessel segments as the left main trunk.

14. The method according to claim 13, wherein Based on the left coronary artery centerline, identify the multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among the multiple vessel segments, and obtain the respective names corresponding to the multiple left coronary artery vessel segments, further including: Determine the left coronary artery vessel segment with the longest length and the left coronary artery vessel segment with the second longest length among the multiple left coronary artery vessel segments, and name the left coronary artery vessel segment closer to the left ventricle among the two determined left coronary artery vessel segments as the anterior descending branch segment, and name the left coronary artery vessel segment closer to the left atrium as the circumflex branch segment.

15. The method according to claim 14, characterized in that Based on the left coronary artery centerline, identify the multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among the multiple vessel segments, and obtain the respective names corresponding to the multiple left coronary artery vessel segments, further including: Determine the two left coronary artery vessel segments that point to the left coronary artery vessel segment named the anterior descending branch segment among the multiple left coronary artery vessel segments, and screen out the first diagonal branch close to the left main trunk and the second diagonal branch far from the left main trunk from the two determined left coronary artery vessel segments; Taking the first diagonal branch and the second diagonal branch as boundaries, divide the left coronary artery vessel segment named the anterior descending branch segment into three segments, and name them as the proximal anterior descending branch, the middle anterior descending branch, and the distal anterior descending branch respectively based on the distance from the left main trunk.

16. The method according to claim 14, characterized in that Based on the left coronary artery centerline, identify the multiple left coronary artery vessel segments corresponding to the left coronary artery centerline among the multiple vessel segments, and obtain the respective names corresponding to the multiple left coronary artery vessel segments, further including: Determine the two left coronary artery vessel segments that point to the left coronary artery vessel segment named the circumflex branch segment among the multiple left coronary artery vessel segments, and screen out the first obtuse marginal branch close to the left main trunk and the second obtuse marginal branch far from the left main trunk from the two determined left coronary artery vessel segments; Taking the first obtuse marginal branch and the second obtuse marginal branch as boundaries, divide the left coronary artery vessel segment named the circumflex branch segment into two segments, and name them as the proximal circumflex branch and the mid-distal circumflex branch respectively based on the distance from the left main trunk.

17. The method according to claim 16, characterized in that Based on the left coronary artery centerline, identifying a plurality of left coronary artery segments corresponding to the left coronary artery centerline among the plurality of blood vessel segments, and obtaining the names corresponding to the plurality of left coronary artery segments respectively, further comprising: Taking the left coronary sinus as the starting pixel point, determining all target blood vessel segments after the left coronary artery segments located in the mid-distal segment of the circumflex branch among the plurality of left coronary artery segments; Naming the target blood vessel segment closest to the left coronary sinus among all the target blood vessel segments as the posterior descending branch after the origin of the circumflex branch, and naming the target blood vessel segment with the longest length as the lateral branch after the origin of the circumflex branch.

18. The method according to claim 1, characterized in that, The segmenting the coronary blood vessels in the coronary blood vessel image based on the coronary artery centerline to obtain a plurality of blood vessel segments includes: Segmenting the coronary blood vessels from the coronary blood vessel image; Based on the coronary artery centerline, segmenting the coronary blood vessels to obtain a plurality of blood vessel segments.

19. A coronary artery blood vessel recognition device, characterized in that, Comprising: A coronary artery centerline extraction module, configured to obtain a coronary blood vessel image and extract a coronary artery centerline from the coronary blood vessel image; A right coronary artery centerline detection module, configured to segment the coronary blood vessels in the coronary blood vessel image based on the coronary artery centerline to obtain a plurality of blood vessel segments, and detect a left coronary artery centerline and a right coronary artery centerline in the coronary artery centerline; A left coronary artery segment name obtaining module, configured to identify a plurality of left coronary artery segments corresponding to the left coronary artery centerline among the plurality of blood vessel segments based on the left coronary artery centerline, and obtain the names corresponding to the plurality of left coronary artery segments respectively; A right coronary artery segment name obtaining module, configured to identify a plurality of right coronary artery segments corresponding to the right coronary artery centerline among the plurality of blood vessel segments based on the right coronary artery centerline, and obtain the names corresponding to the plurality of right coronary artery segments respectively.

20. An electronic device, characterized in that, Comprising: At least one processor; And A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the coronary blood vessel identification method according to any one of claims 1-18.

21. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions for causing a processor to implement the coronary blood vessel identification method according to any one of claims 1-18 when executed.

Images