Method, device and equipment for extracting coronary vessel center line and storage medium

Abstract

The application belongs to the technical field of medical images, and discloses a coronary vessel center line extraction method, device, equipment and storage medium. The method comprises the following steps: acquiring a coronary segmentation image and an aorta segmentation image, and detecting the coronary segmentation image to obtain a first left coronary starting point and a first right coronary starting point; performing inflation processing on the aorta segmentation image to obtain a target aorta segmentation image; performing intersection segmentation on the coronary segmentation image based on the target aorta segmentation image to obtain a final coronary segmentation image; detecting the final coronary segmentation image, and determining a left coronary end point and a right coronary end point after detecting a second left coronary starting point and a second right coronary starting point; and determining a left coronary vessel center line and a right coronary vessel center line based on the first left coronary starting point, the first right coronary starting point, the left coronary end point and the right coronary end point. In this way, the coronary vessel center line can be accurately extracted when the coronary vessel origin is abnormal.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and in particular to a method, apparatus, device, and storage medium for extracting the center line of coronary arteries. Background Technology

[0002] Cardiovascular disease has become one of the major threats to human life. Doctors use automated coronary artery reconstruction technology to assist in the diagnosis of various vascular diseases, but obtaining these quantitative results requires the availability of the vascular centerline. Conventional methods for extracting the vascular centerline are only accurate when the cardiac anatomy is normal, i.e., the origin of the coronary arteries is normal. If the origin of the coronary arteries is abnormal, continuing to use conventional methods for extracting the vascular centerline will not yield accurate results, thus affecting the efficiency of clinical diagnosis. Summary of the Invention

[0003] The main objective of this invention is to provide a method, apparatus, device, and storage medium for extracting the coronary artery centerline, aiming to solve the technical problem in the prior art that the coronary artery centerline cannot be accurately extracted when the coronary artery origin is abnormal.

[0004] To achieve the above objectives, the present invention provides a method for extracting the centerline of coronary arteries, the method comprising the following steps:

[0005] Acquire coronary artery segmentation images and aortic segmentation images, and perform detection on the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point;

[0006] The aortic segmentation image is dilated to obtain the target aortic segmentation image;

[0007] Based on the target aortic segmentation image, the coronary artery segmentation image is subjected to intersection segmentation to obtain the final coronary artery segmentation image;

[0008] The final coronary artery segmentation image is detected, and after the second left coronary artery start point and the second right coronary artery start point are detected, the left coronary artery end point and the right coronary artery end point are determined.

[0009] Based on the first left coronary initiation point, the first right coronary initiation point, the left coronary end point, and the right coronary end point, a centerline extraction algorithm is used to determine the centerline of the left coronary vessels and the centerline of the right coronary vessels.

[0010] Optionally, the step of dilating the aorta in the aortic segmentation image to obtain the target aortic segmentation image includes:

[0011] S21. Determine the dilation rate, and perform dilation processing on the aorta in the aorta segmentation image based on the dilation rate to obtain a new aorta segmentation image;

[0012] S22. Based on the new aortic segmentation image, perform intersection segmentation on the coronary artery segmentation image to obtain the target coronary artery segmentation image;

[0013] S23. Determine whether there are two connected components in the target coronary artery segmentation image. If not, add n units to the dilation rate and use it as the new dilation rate. Repeat steps S21-S23 until there are two connected components in the target coronary artery segmentation image, and use the current new aortic segmentation image as the target aortic segmentation image.

[0014] Optionally, the step of detecting the final coronary artery segmentation image, and determining the left and right coronary artery termination points after detecting the second left coronary artery start point and the second right coronary artery start point, includes:

[0015] The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point;

[0016] Taking the second left coronary artery initiation point and the second right coronary artery initiation point as starting points respectively, the left coronary artery initiation point and the right coronary artery initiation point are determined by the front wave conduction strategy.

[0017] Optionally, the step of detecting the final coronary artery segmentation image, and determining the left and right coronary artery termination points after detecting the second left coronary artery start point and the second right coronary artery start point, includes:

[0018] The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point;

[0019] Based on the final coronary artery segmentation image, the coronary artery endpoints are determined using a machine learning model.

[0020] The terminal point of the coronary artery that is in the same connected region as the starting point of the second left coronary artery is taken as the left coronary artery terminal point, and the terminal point of the coronary artery that is in the same connected region as the starting point of the second right coronary artery is taken as the right coronary artery terminal point.

[0021] Optionally, the step of determining the centerline of the left and right coronary vessels using a centerline extraction algorithm based on the first left coronary initiation point, the first right coronary initiation point, the left coronary end point, and the right coronary end point includes:

[0022] After pairing the first left coronary artery start point with the left coronary artery end point, and the first right coronary artery start point with the right coronary artery end point, the center lines of the left and right coronary arteries are determined using the shortest path algorithm based on the coronary artery centerline intensity map.

[0023] Optionally, determining the centerlines of the left and right coronary vessels based on the coronary artery centerline intensity map using a shortest path algorithm includes:

[0024] Based on the coronary artery centerline intensity map, a first shortest path is generated connecting the starting point of the first left coronary artery and the ending point of the left coronary artery, and a second shortest path is generated connecting the starting point of the first right coronary artery and the ending point of the right coronary artery.

[0025] The centerlines of the left and right coronary vessels are determined based on the first and second shortest paths.

[0026] Optionally, the extraction method further includes:

[0027] Based on the coronary artery segmentation image, a coronary artery centerline intensity map is obtained through a machine learning model or image processing algorithm.

[0028] Furthermore, to achieve the above objectives, the present invention also proposes a device for extracting the coronary artery centerline, the device comprising:

[0029] The acquisition module is used to acquire coronary artery segmentation images and aortic segmentation images, and to detect the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point;

[0030] The acquisition module is also used to perform dilation processing on the aortic segmentation image to obtain a target aortic segmentation image;

[0031] The acquisition module is further configured to perform intersection segmentation on the coronary artery segmentation image based on the target aortic segmentation image to obtain the final coronary artery segmentation image;

[0032] The determination module is used to detect the final coronary artery segmentation image, and after detecting the second left coronary artery start point and the second right coronary artery start point, determine the left coronary artery end point and the right coronary artery end point;

[0033] The determining module is further configured to determine the centerline of the left coronary artery and the centerline of the right coronary artery based on the first left coronary starting point, the first right coronary starting point, the left coronary ending point, and the right coronary ending point, using a centerline extraction algorithm.

[0034] Furthermore, to achieve the above objectives, the present invention also proposes a coronary artery centerline extraction device, which includes: a memory, a processor, and a coronary artery centerline extraction program stored in the memory and executable on the processor. The coronary artery centerline extraction program is configured to implement the steps of the coronary artery centerline extraction method described above.

[0035] In addition, to achieve the above objectives, the present invention also proposes a storage medium storing a coronary artery centerline extraction program, wherein when the coronary artery centerline extraction program is executed by a processor, the steps of the coronary artery centerline extraction method described above are implemented.

[0036] The proposed method for extracting the coronary artery centerline involves acquiring coronary artery segmentation images and aortic segmentation images, and detecting the coronary artery segmentation images to obtain a first left coronary artery origin point and a first right coronary artery origin point. The aortic segmentation image is then dilated to obtain a target aortic segmentation image. Based on the target aortic segmentation image, the coronary artery segmentation images are subjected to intersection segmentation to obtain a final coronary artery segmentation image. The final coronary artery segmentation image is then detected, and after detecting a second left coronary artery origin point and a second right coronary artery origin point, the left coronary artery termination point and the right coronary artery termination point are determined. Based on the first left coronary artery origin point, the first right coronary artery origin point, the left coronary artery termination point, and the right coronary artery termination point, the centerline of the left and right coronary arteries are determined. This method enables precise pairing of the coronary artery origin points when there are abnormalities in the origin of the coronary arteries. By performing intersection segmentation on the coronary artery segmentation images based on the aortic segmentation images, and then pairing the new origin points with the termination points, accurate pairing can be achieved, thereby effectively improving the accuracy of the extracted coronary artery centerline. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the structure of the coronary artery centerline extraction device in the hardware operating environment involved in the embodiments of the present invention;

[0038] Figure 2 This is a flowchart illustrating the first embodiment of the method for extracting the coronary artery centerline of the present invention;

[0039] Figure 3 This is a segmented image of the coronary artery when there is an abnormal origin of the coronary artery, as described in the method for extracting the coronary artery centerline of the present invention.

[0040] Figure 4 This is a segmented image of the coronary artery overlaid with the aorta in the method for extracting the coronary artery centerline of the present invention;

[0041] Figure 5 This is a schematic diagram of the structure of the first deep convolutional neural network in the coronary artery centerline extraction method of the present invention;

[0042] Figure 6 This is a schematic diagram of the image segmentation process performed on the coronary artery segmentation image in the method for extracting the coronary artery centerline of the present invention.

[0043] Figure 7This is a schematic diagram of the final coronary artery segmentation image obtained based on coronary artery segmentation images and aortic segmentation images in the method for extracting the coronary artery centerline of the present invention.

[0044] Figure 8 This is a diagram of the coronary artery centerline in the method for extracting the coronary artery centerline of the present invention;

[0045] Figure 9 This is a flowchart illustrating the second embodiment of the method for extracting the coronary artery centerline of the present invention;

[0046] Figure 10 This is a structural block diagram of the first embodiment of the coronary artery centerline extraction device of the present invention.

[0047] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0048] It should be understood that the specific embodiments described herein are for illustrative purposes only and are not intended to limit the scope of the invention.

[0049] Reference Figure 1 , Figure 1 This is a schematic diagram of the coronary artery centerline extraction device in the hardware operating environment of the embodiment of the present invention.

[0050] like Figure 1 As shown, the coronary artery centerline extraction device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen and an input unit such as a keyboard; optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wireless-Fidelity (Wi-Fi) interface). The memory 1005 may be a high-speed random access memory (RAM) or a stable non-volatile memory (NVM), such as a disk drive. The memory 1005 may also optionally be a storage device independent of the aforementioned processor 1001.

[0051] Those skilled in the art will understand that Figure 1The structure shown does not constitute a limitation on the device for extracting the coronary artery centerline and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0052] like Figure 1 As shown, the memory 1005, which serves as a storage medium, may include an operating system, a network communication module, a user interface module, and a program for extracting the coronary artery centerline.

[0053] exist Figure 1 In the coronary artery centerline extraction device shown, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and memory 1005 in the coronary artery centerline extraction device of the present invention can be set in the coronary artery centerline extraction device, and the coronary artery centerline extraction device calls the coronary artery centerline extraction program stored in the memory 1005 through the processor 1001 and executes the coronary artery centerline extraction method provided in the embodiment of the present invention.

[0054] This invention provides a method for extracting the centerline of coronary arteries, referring to... Figure 2 , Figure 2 This is a flowchart illustrating the first embodiment of a method for extracting the centerline of coronary arteries according to the present invention.

[0055] In this embodiment, the method for extracting the coronary artery centerline includes the following steps:

[0056] Step S10: Acquire coronary artery segmentation images and aortic segmentation images, and detect the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point.

[0057] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a mobile phone, tablet computer, or personal computer, or an electronic device or coronary artery centerline extraction device capable of performing the above functions. The following description uses the coronary artery centerline extraction device as an example to illustrate this embodiment and the subsequent embodiments.

[0058] It should be noted that, as Figure 3The coronary artery segmentation image shown is an image obtained after segmenting the coronary artery image. A coronary artery segmentation image can be understood as an image in which only the coronary arteries are in the foreground, and the rest of the image is the background. The coronary artery segmentation image contains all the coronary arteries in the coronary artery image. The coronary artery image can be a CT (Computed Tomography) image, a two-dimensional coronary angiography image, or an MRI (Magnetic Resonance Imaging) image.

[0059] It should be noted that, as Figure 4 The coronary artery segmentation image shown is an image obtained by performing aortic segmentation on the coronary artery image. The aortic segmentation image can be understood as an image in which only the aorta is in the foreground and the rest of the image is the background. The coronary arteries originate from the aorta and are connected to the aorta through the first left coronary artery origin point and the first right coronary artery origin point.

[0060] In specific implementation, it can be done through methods such as Figure 5 The deep convolutional neural network shown is used to perform coronary artery segmentation on the coronary artery image, thereby obtaining a coronary artery segmentation image. Figure 5 The deep convolutional neural network shown is a pre-trained deep convolutional neural network. Several manually labeled coronary artery images are input into the initial deep convolutional neural network to obtain several predicted coronary artery segmentation images. The manually labeled images are continuously compared with the predicted coronary artery segmentation images, and the differences are continuously fed back to the initial deep convolutional neural network to iteratively update the deep neural network, resulting in a pre-trained deep convolutional neural network, i.e., the first deep convolutional neural network. The first deep convolutional neural network can be used to perform coronary artery segmentation processing on the coronary artery images to obtain coronary artery segmentation images; alternatively, coronary artery segmentation processing can be performed on the coronary artery images using methods such as Frangi filtering to obtain coronary artery segmentation images.

[0061] It should be noted that the first left coronary initiation point refers to the entry point where the coronary artery forms the left coronary artery, and the first right coronary initiation point refers to the entry point where the coronary artery forms the right coronary artery. The coronary arteries mainly include the left coronary artery and the right coronary artery. The coronary arteries can be regarded as a whole coronary artery tree. Then the left coronary artery can be regarded as the left coronary artery tree in the coronary artery tree, and the right coronary artery can be regarded as the right coronary artery tree in the coronary artery tree. It can be understood that the left coronary artery contains the left main coronary artery and the left branch coronary artery, and the right coronary artery contains the right main coronary artery and the right branch coronary artery.

[0062] In a specific implementation, a pre-trained deep convolutional neural network can be used to detect the entry endpoints of coronary artery segmentation images, thereby obtaining the first left coronary artery starting point and the first right coronary artery starting point. The training process of the pre-trained deep convolutional neural network is as follows: several coronary artery segmentation images with manually labeled starting points are input into the initial deep convolutional neural network to obtain the predicted coronary artery starting points. The manually labeled starting points are continuously compared with the predicted starting points, and the differences obtained from the comparison are continuously fed back to the initial deep convolutional neural network, thereby achieving iterative updates of the deep neural network, resulting in a pre-trained deep convolutional neural network, namely the second deep convolutional neural network. The starting points in the coronary artery segmentation images can be detected by the second deep convolutional neural network.

[0063] It should be noted that when a patient's coronary artery origin is abnormal, the origin of the first left coronary artery and the origin of the first right coronary artery will be very close, such as... Figure 4 as well as Figure 6 As shown, the starting points of the left and right coronary arteries are almost connected at the same point. The first left coronary artery starting point and the first right coronary artery starting point correspond to... Figure 6 P1 and P2 in the middle can be obtained from Figure 6 As seen in the image on the left, the left coronary artery endpoint and the right coronary artery endpoint are in the same connected region. This may cause the first left coronary artery starting point to match the right coronary artery endpoint, or the first right coronary artery starting point to match the left coronary artery endpoint, thus resulting in inaccurate extraction of the coronary artery centerline.

[0064] Step S20: Dilate the aorta in the aortic segmentation image to obtain the target aortic segmentation image.

[0065] It should be noted that the expansion process can be understood as proportional scaling.

[0066] It is understandable that the target aortic segmentation image is the same size as the aortic segmentation image. After dilating the aorta in the aortic segmentation image, the foreground part of the aortic segmentation image, i.e. the aorta, will be enlarged proportionally to obtain the target aortic segmentation image. The foreground part in the target aortic segmentation image is the aorta enlarged proportionally.

[0067] Step S30: Based on the target aortic segmentation image, perform intersection segmentation on the coronary artery segmentation image to obtain the final coronary artery segmentation image.

[0068] It should be noted that performing intersection segmentation on the coronary artery segmentation image based on the target aorta segmentation image can be understood as overlaying the foreground portion (i.e., the dilated aorta) in the target aorta segmentation image onto the corresponding aorta location in the coronary artery segmentation image, and then converting the overlaid area in the coronary artery segmentation image into the background portion of the image, thus obtaining the final coronary artery segmentation image. For example... Figure 5 As shown, this is the image obtained by overlaying the foreground portion of the target aorta segmentation image (i.e., the aorta after dilation) onto the aorta position in the coronary artery segmentation image.

[0069] Understandably, the final coronary artery segmentation image can be interpreted as a small portion of the coronary artery initiation image being truncated, thus allowing the starting points of the left and right coronary arteries to be clearly seen in the final image. Figure 6 As shown in the right image, two connected components are formed, so that the left crown endpoint and the right crown endpoint are no longer in the same connected component.

[0070] It should be noted that, as Figure 7 As shown, Figure 7 The leftmost image is the coronary artery segmentation image, the middle image is the aortic segmentation image, and the rightmost image is the final coronary artery segmentation image. From the final coronary artery segmentation image, it can be seen that the starting points of the left and right coronary arteries in the segmented coronary arteries are separated by a certain distance.

[0071] Understandably, the final coronary artery segmentation image is only missing a small segment of the coronary artery at the beginning of the coronary artery compared to the original coronary artery segmentation image.

[0072] Step S40: Detect the final coronary artery segmentation image. After detecting the second left coronary artery start point and the second right coronary artery start point, determine the left coronary artery end point and the right coronary artery end point.

[0073] Understandably, the left coronary artery contains the left main coronary artery and several left branch coronary arteries, and the right coronary artery contains the right main coronary artery and several right branch coronary arteries. Each main coronary artery and each branch coronary artery consists of a starting point and an ending point. The left main coronary artery and several left branch coronary arteries all originate from the first left coronary artery starting point, and the right main coronary artery and several right branch coronary arteries all originate from the first right coronary artery starting point. The ending points of each main coronary artery and each branch coronary artery are different. Therefore, a coronary artery contains multiple ending points. The number of ending points of the left coronary artery is not limited to one, and the number of ending points of the right coronary artery is not limited to one either.

[0074] It should be noted that the second left coronary artery starting point and the second right coronary artery starting point obtained by entrance detection in the final coronary artery segmentation image are spaced a certain distance apart, so that there will be no confusion in matching the left coronary artery ending point and the right coronary artery ending point due to the first left coronary artery starting point and the first right coronary artery starting point being too close; therefore, the second left coronary artery starting point and the second right coronary artery starting point can accurately match the left coronary artery ending point and the right coronary artery ending point.

[0075] In specific implementations, such as Figure 6 As shown, Figure 6 P3 and P4 correspond to the starting points of the second left and second right coronary arteries, respectively. Figure 6 It can be determined that after the coronary artery segmentation image is segmented by intersection based on the target aortic segmentation image, two connected regions are formed, so that the left coronary artery terminal point and the right coronary artery terminal point are no longer in the same connected region.

[0076] In a specific implementation, the entry endpoints of the final coronary artery segmentation image can be detected using a deep convolutional neural network, thereby obtaining the second left coronary artery starting point and the second right coronary artery starting point. Specifically, the entry endpoints of the final coronary artery segmentation image can be detected using a second deep convolutional neural network.

[0077] In one embodiment, the step of detecting the final coronary artery segmentation image, and determining the left and right coronary artery termination points after detecting the second left coronary artery start point and the second right coronary artery start point, includes:

[0078] The final coronary artery segmentation image is detected to obtain the second left coronary artery start point and the second right coronary artery start point; based on the final coronary artery segmentation image, the coronary artery end points are determined; the end point of the coronary artery end points that is in the same connected region as the second left coronary artery start point is taken as the left coronary artery end point, and the end point of the coronary artery end points that is in the same connected region as the second right coronary artery start point is taken as the right coronary artery end point.

[0079] It should be noted that since the coronary arteries include the main coronary artery and several branch coronary arteries, it can be determined that the final coronary artery segmentation image contains multiple coronary artery terminal points. Due to the large number of coronary artery terminal points, it is not possible to directly determine the left and right coronary artery terminal points from among the multiple terminal points. Therefore, the terminal points that are in the same connected region as the second left coronary artery starting point can be identified as the left terminal points, and the terminal points that are in the same connected region as the second right coronary artery starting point can be identified as the right terminal points.

[0080] In a practical implementation, a deep convolutional network can be used to determine the endpoints that are in the same connected domain as the starting point of the second left crown and the endpoints that are in the same connected domain as the starting point of the second right crown.

[0081] In one embodiment, the step of detecting the final coronary artery segmentation image, and determining the left and right coronary artery termination points after detecting the second left coronary artery start point and the second right coronary artery start point, includes:

[0082] The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point; the coronary artery termination point is determined based on the final coronary artery segmentation image; the left coronary artery termination point and the right coronary artery termination point are determined from the second left coronary artery starting point and the second right coronary artery starting point, respectively, using the forward wave propagation strategy.

[0083] It should be noted that the forward wave propagation strategy refers to the region growing algorithm. The region growing algorithm can quickly and accurately identify the terminal points that are in the same connected domain as the second left coronary artery initiation point, and can also quickly and accurately identify the terminal points that are in the same connected domain as the second right coronary artery initiation point, thereby obtaining the left and right coronary artery terminal points among the coronary artery terminal points.

[0084] Step S50: Based on the first left coronary starting point, the first right coronary starting point, the left coronary ending point, and the right coronary ending point, the centerline extraction method is used to determine the centerline of the left coronary vessel and the centerline of the right coronary vessel.

[0085] It should be noted that the vascular centerline can be understood as the centerline connecting the head and tail of a blood vessel. The left coronary artery centerline refers to the centerline of the entire left coronary artery in the coronary image, and the right coronary artery centerline refers to the centerline of the entire right coronary artery in the coronary image. Specifically, as shown below... Figure 7 The diagram shows the centerline of the coronary arteries.

[0086] In the specific implementation, after pairing the first left coronary starting point with the left coronary ending point, and pairing the first right coronary ending point with the right coronary ending point, the centerline of the left coronary vessel and the centerline of the right coronary vessel can be determined by a learning-based path tracing algorithm.

[0087] In one embodiment, the step of determining the centerline of the left and right coronary vessels using a centerline extraction method based on the first left coronary initiation point, the first right coronary initiation point, the left coronary end point, and the right coronary end point includes:

[0088] After pairing the first left coronary artery start point with the left coronary artery end point, and the first right coronary artery start point with the right coronary artery end point, the center lines of the left and right coronary arteries are determined using the shortest path algorithm based on the coronary artery centerline intensity map.

[0089] It should be noted that the coronary artery centerline intensity map refers to the probability that each pixel in the coronary artery segmentation image is the coronary artery centerline.

[0090] In the specific implementation, the first left coronary artery starting point is paired one-to-one with all left coronary artery ending points, and the first right coronary artery starting point is paired one-to-one with all right coronary artery ending points, resulting in several matching pairs. Each matching pair consists of a starting point and an ending point. The shortest path algorithm can be used to generate the shortest path for each matching pair, thereby determining the centerline of the left and right coronary arteries. The shortest path algorithm can be Dijkstra's algorithm, fast marching, or other algorithms.

[0091] In one embodiment, determining the centerlines of the left and right coronary vessels based on the coronary centerline intensity map using a shortest path algorithm includes:

[0092] Based on the coronary artery centerline intensity map, a first shortest path is generated connecting the starting point of the first left coronary artery and the ending point of the left coronary artery, and a second shortest path is generated connecting the starting point of the first right coronary artery and the ending point of the right coronary artery.

[0093] The centerlines of the left and right coronary vessels are determined based on the first and second shortest paths.

[0094] It should be noted that the coronary artery centerline intensity map refers to the probability that each pixel in the coronary artery segmentation image is the coronary artery centerline.

[0095] In the specific implementation, after pairing the first left coronary starting point with all left coronary ending points one by one, several left matching pairs are obtained. Similarly, after pairing the first right coronary starting point with all right coronary ending points one by one, several right matching pairs are obtained. Each left matching pair consists of a left starting point and a left ending point, and each right matching pair consists of a right starting point and a right ending point. The shortest path algorithm can be used to generate the shortest path for several left matching pairs, i.e., the first shortest path, and the shortest path algorithm can be used to generate the shortest path for several right matching pairs, i.e., the second shortest path. The center lines of the left and right coronary vessels are then determined based on the first and second shortest paths, respectively.

[0096] In one embodiment, the extraction method further includes:

[0097] Based on the coronary artery segmentation image, a coronary artery centerline intensity map is obtained through a machine learning model or image processing algorithm.

[0098] This embodiment acquires coronary artery segmentation images and aortic segmentation images, and detects the coronary artery segmentation images to obtain the first left coronary artery origin point and the first right coronary artery origin point. The aortic segmentation image is then dilated to obtain a target aortic segmentation image. Based on the target aortic segmentation image, the coronary artery segmentation images are intersected to obtain the final coronary artery segmentation image. The final coronary artery segmentation image is then detected, and after detecting the second left coronary artery origin point and the second right coronary artery origin point, the left coronary artery termination point and the right coronary artery termination point are determined. Based on the first left coronary artery origin point, the first right coronary artery origin point, the left coronary artery termination point, and the right coronary artery termination point, a centerline extraction method is used to determine the centerline of the left and right coronary arteries. Through this method, when the origin of the coronary arteries is abnormal, a new coronary artery origin point can be determined after intersecting the coronary artery segmentation images based on the aortic segmentation image. This new coronary artery origin point can then be used to pair with the coronary artery termination points, achieving accurate pairing and effectively improving the accuracy of the extracted coronary artery centerline.

[0099] refer to Figure 9 , Figure 9 This is a flowchart illustrating a second embodiment of a method for extracting the centerline of a coronary artery according to the present invention.

[0100] Based on the first embodiment described above, step S20 in the coronary artery centerline extraction method of this embodiment includes:

[0101] Step S21: Determine the dilation rate, and perform dilation processing on the aorta in the aorta segmentation image based on the dilation rate to obtain a new aorta segmentation image.

[0102] Step S22: Based on the new aortic segmentation image, perform intersection segmentation on the coronary artery segmentation image to obtain the target coronary artery segmentation image.

[0103] Step S23: Determine whether there are two connected components in the target coronary artery segmentation image. If not, add n units to the dilation rate and use it as the new dilation rate. Repeat steps S21-S23 until there are two connected components in the target coronary artery segmentation image, and use the current new aortic segmentation image as the target aortic segmentation image.

[0104] It should be noted that the dilation rate can be understood as a proportional magnification factor; the intersection segmentation of the coronary artery segmentation image based on the new aortic segmentation image can be understood as covering the foreground part of the new aortic segmentation image onto the aortic position in the coronary artery segmentation image, and then converting the covered area in the coronary artery segmentation image into the background part of the image, so as to obtain the target coronary artery segmentation image.

[0105] Understandably, when the expansion rate is too large, the coronary artery segmentation image will be incomplete after the intersection segmentation of the new aortic segmentation image and the coronary artery segmentation image. For example, because the foreground part in the new aortic segmentation image is too large, when the foreground part of the new aortic segmentation image is covered on the coronary artery segmentation image, it may cover multiple coronary artery branches in the coronary artery segmentation image, resulting in the incomplete coronary artery centerline extracted in the end.

[0106] It should be noted that the expansion rate can be adjusted by increasing the expansion rate by n units in each iteration. Specifically, n can be set according to the actual situation to prevent the expansion rate from becoming too large after each increase of n units.

[0107] This embodiment continuously adjusts the dilation rate of the aortic segmentation image to determine the optimal dilation rate, enabling intersection segmentation without covering other coronary artery branches. This ensures that there are two relatively clear connected regions in the target coronary artery segmentation image, thereby enabling the rapid and accurate pairing of the first left coronary artery start point with the left coronary artery end point and the first right coronary artery start point with the right coronary artery end point, thus improving the accuracy of the final extracted coronary artery centerline.

[0108] Furthermore, this embodiment of the invention also proposes a storage medium storing a coronary artery centerline extraction program, which, when executed by a processor, implements the steps of the coronary artery centerline extraction method described above.

[0109] Reference Figure 10 , Figure 10 This is a structural block diagram of the first embodiment of the coronary artery centerline extraction device of the present invention.

[0110] like Figure 10 As shown, the coronary artery centerline extraction device proposed in this embodiment of the invention includes:

[0111] The acquisition module 10 is used to acquire coronary artery segmentation images and aortic segmentation images, and to detect the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point.

[0112] The acquisition module 10 is further configured to perform dilation processing on the aortic segmentation image to obtain a target aortic segmentation image.

[0113] The acquisition module 10 is further configured to perform intersection segmentation on the coronary artery segmentation image based on the target aortic segmentation image to obtain the final coronary artery segmentation image.

[0114] The determination module 20 is used to detect the final coronary artery segmentation image and, after detecting the second left coronary artery starting point and the second right coronary artery starting point, determine the left coronary artery ending point and the right coronary artery ending point.

[0115] The determining module 20 is further configured to determine the centerline of the left coronary artery and the centerline of the right coronary artery based on the first left coronary starting point, the first right coronary starting point, the left coronary ending point, and the right coronary ending point, using a centerline extraction algorithm.

[0116] It should be understood that the above are merely illustrative examples and do not constitute any limitation on the technical solutions of the present invention. In specific applications, those skilled in the art can make settings as needed, and the present invention does not impose any restrictions on this.

[0117] This embodiment acquires coronary artery segmentation images and aortic segmentation images, and detects the coronary artery segmentation images to obtain the first left coronary artery origin point and the first right coronary artery origin point. The aortic segmentation image is then dilated to obtain a target aortic segmentation image. Based on the target aortic segmentation image, the coronary artery segmentation images are intersected to obtain the final coronary artery segmentation image. The final coronary artery segmentation image is then detected, and after detecting the second left coronary artery origin point and the second right coronary artery origin point, the left coronary artery termination point and the right coronary artery termination point are determined. Based on the first left coronary artery origin point, the first right coronary artery origin point, the left coronary artery termination point, and the right coronary artery termination point, a centerline extraction method is used to determine the centerline of the left and right coronary arteries. Through this method, when the origin of the coronary arteries is abnormal, a new coronary artery origin point can be determined after intersecting the coronary artery segmentation images based on the aortic segmentation image. This new coronary artery origin point can then be used to pair with the coronary artery termination points, achieving accurate pairing and effectively improving the accuracy of the extracted coronary artery centerline.

[0118] In one embodiment, the acquisition module 10 is further configured to:

[0119] Determine the dilation rate, and dilate the aorta in the aortic segmentation image based on the dilation rate to obtain a new aortic segmentation image;

[0120] Based on the new aortic segmentation image, the coronary artery segmentation image is subjected to intersection segmentation to obtain the target coronary artery segmentation image;

[0121] Determine whether there are two connected components in the target coronary artery segmentation image. If not, increase the dilation rate by n units and use it as the new dilation rate. Repeat steps S21-S23 until there are two connected components in the target coronary artery segmentation image, and use the current new aortic segmentation image as the target aortic segmentation image.

[0122] In one embodiment, the determining module 20 is further configured to:

[0123] The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point;

[0124] Taking the second left coronary artery initiation point and the second right coronary artery initiation point as starting points respectively, the left coronary artery initiation point and the right coronary artery initiation point are determined by the front wave conduction strategy.

[0125] In one embodiment, the determining module 20 is further configured to:

[0126] The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point;

[0127] Based on the final coronary artery segmentation image, the coronary artery endpoints are determined using a machine learning model.

[0128] The terminal point of the coronary artery that is in the same connected region as the starting point of the second left coronary artery is taken as the left coronary artery terminal point, and the terminal point of the coronary artery that is in the same connected region as the starting point of the second right coronary artery is taken as the right coronary artery terminal point.

[0129] In one embodiment, the determining module 20 is further configured to:

[0130] After pairing the first left coronary artery start point with the left coronary artery end point, and the first right coronary artery start point with the right coronary artery end point, the center lines of the left and right coronary arteries are determined using the shortest path algorithm based on the coronary artery centerline intensity map.

[0131] In one embodiment, the determining module 20 is further configured to:

[0132] Based on the coronary artery centerline intensity map, a first shortest path is generated connecting the starting point of the first left coronary artery and the ending point of the left coronary artery, and a second shortest path is generated connecting the starting point of the first right coronary artery and the ending point of the right coronary artery.

[0133] The centerlines of the left and right coronary vessels are determined based on the first and second shortest paths.

[0134] In one embodiment, the acquisition module 10 is further configured to:

[0135] Based on the coronary artery segmentation image, a coronary artery centerline intensity map is obtained through a machine learning model or image processing algorithm.

[0136] It should be noted that the workflow described above is merely illustrative and does not limit the scope of protection of this invention. In practical applications, those skilled in the art can select some or all of the workflow to achieve the purpose of this embodiment according to actual needs, and no restrictions are imposed here.

[0137] In addition, for technical details not described in detail in this embodiment, please refer to the method for extracting the coronary artery centerline provided in any embodiment of the present invention, which will not be repeated here.

[0138] Furthermore, it should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0139] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0140] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as read-only memory (ROM) / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0141] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for extracting the centerline of a coronary artery, characterized in that, The extraction method includes: Acquire coronary artery segmentation images and aortic segmentation images, and perform detection on the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point; The aortic segmentation image is dilated to obtain the target aortic segmentation image; Based on the target aortic segmentation image, the coronary artery segmentation image is subjected to intersection segmentation to obtain the final coronary artery segmentation image; this step specifically includes: covering the aorta in the target aortic segmentation image with the corresponding aortic position in the coronary artery segmentation image to obtain the intersection portion; removing the intersection portion from the coronary artery segmentation image to obtain the final coronary artery segmentation image; The final coronary artery segmentation image is detected. After the second left coronary artery start point and the second right coronary artery start point are detected, the left coronary artery end point and the right coronary artery end point are determined based on the second left coronary artery start point and the second right coronary artery start point. Based on the first left coronary initiation point, the first right coronary initiation point, the left coronary end point, and the right coronary end point, a centerline extraction algorithm is used to determine the centerline of the left coronary vessels and the centerline of the right coronary vessels.

2. The method as described in claim 1, characterized in that, The step of dilating the aorta in the segmented aorta image to obtain the target aorta segmented image includes: S21. Determine the dilation rate, and perform dilation processing on the aorta in the aorta segmentation image based on the dilation rate to obtain a new aorta segmentation image; S22. Based on the new aortic segmentation image, perform intersection segmentation on the coronary artery segmentation image to obtain the target coronary artery segmentation image; S23. Determine whether there are two connected components in the target coronary artery segmentation image. If not, add n units to the dilation rate and use it as the new dilation rate. Repeat steps S21-S23 until there are two connected components in the target coronary artery segmentation image, and use the current new aortic segmentation image as the target aortic segmentation image.

3. The method as described in claim 1, characterized in that, The step of detecting the final coronary artery segmentation image, after detecting the second left coronary artery start point and the second right coronary artery start point, determines the left coronary artery end point and the right coronary artery end point based on the second left coronary artery start point and the second right coronary artery start point, including: The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point; Starting from the second left coronary artery initiation point and the second right coronary artery initiation point respectively, the left coronary artery terminal point and the right coronary artery terminal point are determined by the front wave propagation strategy; the front wave propagation strategy is a region growing algorithm.

4. The method as described in claim 1, characterized in that, The step of detecting the final coronary artery segmentation image, after detecting the second left coronary artery start point and the second right coronary artery start point, determines the left coronary artery end point and the right coronary artery end point based on the second left coronary artery start point and the second right coronary artery start point, including: The final coronary artery segmentation image is detected to obtain the second left coronary artery starting point and the second right coronary artery starting point; Based on the final coronary artery segmentation image, the coronary artery endpoints are determined using a machine learning model. The terminal point of the coronary artery that is in the same connected region as the starting point of the second left coronary artery is taken as the left coronary artery terminal point, and the terminal point of the coronary artery that is in the same connected region as the starting point of the second right coronary artery is taken as the right coronary artery terminal point.

5. The method as described in claim 1, characterized in that, The determination of the centerlines of the left and right coronary vessels based on the first left coronary initiation point, the first right coronary initiation point, the left coronary end point, and the right coronary end point, using a centerline extraction algorithm, includes: After pairing the first left coronary artery start point with the left coronary artery end point, and the first right coronary artery start point with the right coronary artery end point, the center lines of the left and right coronary arteries are determined using the shortest path algorithm based on the coronary artery centerline intensity map.

6. The method as described in claim 5, characterized in that, The determination of the left and right coronary vessel centerlines based on the coronary artery centerline intensity map using a shortest path algorithm includes: Based on the coronary artery centerline intensity map, a first shortest path is generated connecting the starting point of the first left coronary artery and the ending point of the left coronary artery, and a second shortest path is generated connecting the starting point of the first right coronary artery and the ending point of the right coronary artery. The centerlines of the left and right coronary vessels are determined based on the first and second shortest paths.

7. The method as described in any one of claims 5 or 6, characterized in that, The extraction method further includes: Based on the coronary artery segmentation image, a coronary artery centerline intensity map is obtained through a machine learning model or image processing algorithm.

8. A device for extracting the centerline of a coronary artery, characterized in that, The extraction device includes: The acquisition module is used to acquire coronary artery segmentation images and aortic segmentation images, and to detect the coronary artery segmentation images to obtain the first left coronary artery starting point and the first right coronary artery starting point; The acquisition module is also used to perform dilation processing on the aortic segmentation image to obtain a target aortic segmentation image; The acquisition module is further configured to perform intersection segmentation on the coronary artery segmentation image based on the target aortic segmentation image to obtain a final coronary artery segmentation image; the acquisition module is further configured to overlay the dilated aorta in the target aortic segmentation image onto the aortic position in the coronary artery segmentation image to obtain an intersection portion; and remove the intersection portion from the coronary artery segmentation image to obtain the final coronary artery segmentation image; The determination module is used to detect the final coronary artery segmentation image. After detecting the second left coronary artery start point and the second right coronary artery start point, it determines the left coronary artery end point and the right coronary artery end point based on the second left coronary artery start point and the second right coronary artery start point. The determining module is further configured to determine the centerline of the left coronary artery and the centerline of the right coronary artery based on the first left coronary starting point, the first right coronary starting point, the left coronary ending point, and the right coronary ending point, using a centerline extraction algorithm.

9. A device for extracting the centerline of coronary arteries, characterized in that, The device includes: a memory, a processor, and a coronary artery centerline extraction program stored in the memory and executable on the processor, the coronary artery centerline extraction program being configured to implement the steps of the coronary artery centerline extraction method as described in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium stores a coronary artery centerline extraction program, which, when executed by a processor, implements the steps of the coronary artery centerline extraction method as described in any one of claims 1 to 7.

Images