A coronary CTA processing method and device for vascular occlusion lesions

Abstract

The present invention discloses a coronary CTA processing method and device for vascular occlusion lesions, relating to the field of medical imaging technology. A specific embodiment of the method includes: identifying the endpoints of coronary blood vessels from coronary CTA images; identifying a position to be discriminated that is suspected of having a chronic total occlusion (CTO); wherein the position to be discriminated is a blood vessel without blood flow signal between two adjacent endpoints; using a pre-determined set of CTO detection models to detect the position to be discriminated, and obtaining a CTO discrimination result for the position to be discriminated; wherein the set of CTO detection models contains a plurality of pre-trained CTO detection models corresponding to different CTO scale intervals. This embodiment can improve the CTO detection accuracy of coronary CTA images.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and in particular to a coronary CTA processing method and device for vascular occlusion lesions. Background Art

[0002] CTO (Chronic Total Occlusion) is a serious vascular lesion. It is of great significance to accurately detect whether there is CTO in coronary artery images. During the automatic detection process of coronary CTA (i.e., coronary CT, a coronary artery examination method based on computed tomography CT) images, thinner arterial branches and the ends of the main arterial branches are prone to be confused with veins, resulting in the inability to distinguish whether a certain position without blood flow signal is a CTO or a vein abutment when making a judgment, thus affecting the accuracy of CTO detection. In addition, different scales of CTO often correspond to different overall vascular connection situations. However, in the existing automatic CTO detection process, the same machine learning model is often used to detect CTOs of various scales, resulting in a further decline in the accuracy of CTO detection. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a coronary CTA processing method and device for vascular occlusion lesions, which can improve the accuracy of CTO detection in coronary CTA images.

[0004] To achieve the above object, according to one aspect of the present invention, a coronary CTA processing method for vascular occlusion lesions is provided.

[0005] The coronary CTA processing method for vascular occlusion lesions in the embodiments of the present invention includes: identifying the end points of coronary blood vessels from coronary CTA images; identifying a position to be discriminated suspected of chronic total occlusion CTO; wherein the position to be discriminated is a blood vessel without blood flow signal between two adjacent end points; detecting the position to be discriminated by using a pre-determined set of CTO detection models to obtain a CTO discrimination result of the position to be discriminated; wherein the set of CTO detection models includes multiple pre-trained CTO detection models corresponding to different CTO scale intervals.

[0006] Optionally, the identifying the end points of coronary blood vessels from coronary CTA images includes: inputting the coronary CTA images into a pre-trained vascular segmentation model to obtain the coronary blood vessels in the coronary CTA images; performing skeleton extraction on the coronary blood vessels, and determining the center line of the coronary blood vessels according to the skeleton extraction result; wherein the pixel connection of the center line is a tree structure, and the root node of the tree structure is a pixel of the aorta; traversing each pixel in the tree structure from the root node, and determining the non-root node pixel with only one adjacent pixel as the end point of the coronary blood vessel.

[0007] Optionally, any CTO detection model is trained through the following steps: obtaining CTO lesion images and non-CTO images; wherein, the CTO scale in the CTO lesion images is within a preset CTO scale range for the CTO detection model; using the CTO lesion images and the non-CTO images as training samples, and whether the CTO lesion images or the non-CTO images have CTO as labels to train the CTO detection model.

[0008] Optionally, detecting the to-be-discriminated position by using the pre-determined set of CTO detection models to obtain the CTO discrimination result of the to-be-discriminated position includes: calculating the scale of the to-be-discriminated position, and determining the CTO detection model corresponding to this scale from the set of CTO detection models; segmenting the associated image of the to-be-discriminated position from the coronary CTA image based on the two end points of the to-be-discriminated position, and inputting the associated image into the determined CTO detection model; if the CTO detection model outputs a positive result, using this result as the CTO discrimination result of the to-be-discriminated position.

[0009] Optionally, the coronary CTA image is a three-dimensional image; and, segmenting the associated image of the to-be-discriminated position from the coronary CTA image based on the two end points of the to-be-discriminated position includes: in the coronary CTA image, respectively expanding from the two end points of the to-be-discriminated position in a direction away from the to-be-discriminated position until the three-dimensional area included between the two expanded parts reaches a preset first three-dimensional size; determining the image of the three-dimensional area that reaches the first three-dimensional size as the associated image of the to-be-discriminated position.

[0010] Optionally, detecting the to-be-discriminated position by using the pre-determined set of CTO detection models to obtain the CTO discrimination result of the to-be-discriminated position further includes: if the determined CTO detection model outputs a negative result, respectively shrinking the two end points of the to-be-discriminated position by a preset distance in a direction close to the center of the to-be-discriminated position to obtain two new end points of the to-be-discriminated position; calculating the scale between the two new end points, and querying the CTO detection model corresponding to this scale from the set of CTO detection models; segmenting the corresponding associated image from the coronary CTA image based on the two new end points, inputting the associated image into the queried CTO detection model, and using the output result of this CTO detection model as the CTO discrimination result of the to-be-discriminated position.

[0011] Optionally, segmenting the corresponding associated image from the coronary CTA image based on the two new endpoints includes: in the coronary CTA image, respectively expanding from the two new endpoints in a direction away from the position to be discriminated until the three-dimensional region included between the two expanded parts reaches a preset second three-dimensional size; wherein the second three-dimensional size is smaller than the first three-dimensional size in each dimension; and determining the image of the three-dimensional region that reaches the second three-dimensional size as the associated image corresponding to the two new endpoints.

[0012] To achieve the above object, according to another aspect of the present invention, there is provided a coronary CTA processing device for vascular occlusion lesions.

[0013] The coronary CTA processing device for vascular occlusion lesions according to an embodiment of the present invention may include: a vascular endpoint extraction unit for identifying endpoints of coronary blood vessels from a coronary CTA image; a position to be discriminated acquisition unit for: identifying a position to be discriminated where a chronic total occlusion (CTO) is suspected to occur; wherein the position to be discriminated is a blood vessel without blood flow signal between two adjacent endpoints; a CTO discrimination unit for: detecting the position to be discriminated by using a pre-determined set of CTO detection models to obtain a CTO discrimination result of the position to be discriminated; wherein the set of CTO detection models includes a plurality of pre-trained CTO detection models corresponding to different CTO scale intervals.

[0014] To achieve the above object, according to still another aspect of the present invention, there is provided an electronic device.

[0015] An electronic device according to the present invention includes: one or more processors; a storage device for storing one or more programs, which when executed by the one or more processors cause the one or more processors to implement the method of the coronary CTA processing device for vascular occlusion lesions provided by the present invention.

[0016] To achieve the above object, according to yet another aspect of the present invention, there is provided a computer-readable storage medium.

[0017] A computer-readable storage medium according to the present invention has a computer program stored thereon, and when the program is executed by a processor, it implements the method of the coronary CTA processing device for vascular occlusion lesions provided by the present invention.

[0018] According to the technical solution of the present invention, the embodiments in the above invention have the following advantages or beneficial effects:

[0019] First, identify the endpoints of the coronary arteries from coronary CTA images; then, identify the location to be discriminated that is suspected of having a chronic total occlusion (CTO), where the location to be discriminated is the blood vessel without blood flow signal between two adjacent endpoints; finally, use a CTO detection model set including multiple CTO detection models corresponding to different CTO scale intervals to detect the location to be discriminated, and obtain the CTO discrimination result of the location to be discriminated. Through the above steps, it is possible to use the CTO detection model adapted to the scale of the location to be discriminated in the set for targeted judgment, thereby improving the accuracy of CTO detection. In addition, the existing blood vessel segmentation process is prone to expanding the scale of the location to be discriminated, resulting in the subsequent use of a CTO detection model with a mismatched scale. To address the above defects, after detecting based on the original scale of the location to be discriminated, the present invention shrinks the location to be discriminated according to a preset strategy, determines the corresponding CTO detection model according to the shrunk scale, and performs re-detection, thereby avoiding CTO missed detection and helping to obtain a more accurate CTO detection result.

[0020] The further effects of the above non-conventional optional manner will be described below in conjunction with specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The drawings are used to better understand the present invention and do not constitute an improper limitation to the present invention. Among them:

[0022] Figure 1 is a schematic diagram of the main steps of the coronary CTA processing method for vascular occlusion lesions in an embodiment of the present invention;

[0023] Figure 2 is a schematic diagram of CTO in an embodiment of the present invention;

[0024] Figure 3 is a schematic diagram of the shrinkage of the location to be discriminated in an embodiment of the present invention;

[0025] Figure 4 is a schematic diagram of the specific execution steps of the coronary CTA processing method for vascular occlusion lesions in an embodiment of the present invention;

[0026] Figure 5 is a schematic diagram of the components of the coronary CTA processing device for vascular occlusion lesions in an embodiment of the present invention;

[0027] Figure 6 is an exemplary system architecture diagram to which the embodiment of the present invention can be applied;

[0028] Figure 7 is a schematic diagram of the structure of an electronic device for implementing the coronary CTA processing method for vascular occlusion lesions in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0029] The following describes exemplary embodiments of the present invention with reference to the accompanying drawings. Various details of the embodiments of the present invention are included to facilitate understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present invention. Similarly, descriptions of well-known functions and structures are omitted in the following description for clarity and conciseness.

[0030] It should be noted that, without conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0031] Figure 1 It is a schematic diagram of the main steps of the coronary CTA processing method for vascular occlusion lesions according to the embodiments of the present invention.

[0032] As shown in Figure 1 the following, the coronary CTA processing method for vascular occlusion lesions in the embodiments of the present invention can be specifically executed according to the following steps:

[0033] Step S101: Identify the endpoints of the coronary blood vessels from the coronary CTA images.

[0034] In the embodiments of the present invention, the above coronary CTA images are three-dimensional images. In specific applications, after obtaining the coronary CTA images, the coronary blood vessels and the endpoints of the coronary blood vessels can be identified therefrom to further determine the location where CTO is suspected to occur subsequently. Figure 2 It is a schematic diagram of CTO in the embodiments of the present invention. As shown in Figure 2 the following, the CTO in the blood vessel is indicated by the arrow, and the rectangle represents the aorta.

[0035] As a preferred solution, the endpoints of the coronary blood vessels can be identified in the following way: First, input the coronary CTA images into a pre-trained blood vessel segmentation model to obtain the coronary blood vessels in the coronary CTA images. Among them, the blood vessel segmentation model can be implemented based on known algorithms, and the details of its training and use are not required to be described here. After obtaining the coronary blood vessels, perform skeleton extraction on the coronary blood vessels, and determine the centerline of the coronary blood vessels based on the skeleton extraction results. In particular, the pixels of the centerline can be connected into a tree structure, and the root node of the above tree structure is a pixel of the aorta. Finally, traverse each pixel in the tree structure from the root node. If a pixel is adjacent to two pixels, it is an intermediate point of the coronary blood vessel; if a pixel has only one adjacent pixel and this pixel is not the root node pixel, it is determined as the endpoint of the coronary blood vessel.

[0036] Step S102: Identify the position to be discriminated where chronic total occlusion CTO is suspected to occur.

[0037] In this step, the blood vessel without blood flow signal between the two identified adjacent endpoints can be used as the position to be judged for suspected CTO, and the subsequent steps will perform the judgment on whether CTO has occurred for the position to be judged.

[0038] Step S103: Detect the position to be judged by using the pre-determined CTO detection model set, and obtain the CTO judgment result of the position to be judged.

[0039] In practical applications, the above CTO detection model set includes multiple pre-trained CTO detection models corresponding to different CTO scale intervals. Among them, the CTO scale refers to the CTO occlusion length in the blood vessel. Generally, the above multiple CTO detection models are similar models, all of which contain a convolutional layer for extracting image features and a classification layer for classifying according to the extracted image features. In one embodiment, different CTO detection models may have convolutional layers with different parameters and the same classification layer. Exemplarily, the CTO detection model set includes 7 CTO detection models, corresponding to the following CTO scale intervals respectively: (0 mm, 2 mm], (2 mm, 4 mm], (4 mm, 6 mm], (6 mm, 8 mm], (8 mm, 10 mm], (10 mm, 12 mm], (12 mm, 14 mm].

[0040] Preferably, any CTO detection model can be trained through the following steps: First, obtain CTO lesion images and non-CTO images; wherein, the CTO scale in the CTO lesion images is within the CTO scale interval preset for this CTO detection model; thereafter, use the CTO lesion images and the non-CTO images as training samples, and whether the CTO lesion images or the non-CTO images have CTO as labels to train this CTO detection model. It can be understood that through the above training, each CTO detection model can learn the association between the overall connection situation of the blood vessels in the training samples and the existence of CTO, so as to accurately distinguish coronary CTO from venous abutment, thereby improving the CTO detection accuracy.

[0041] In an embodiment of the present invention, the following steps are specifically performed to achieve CTO detection: First, calculate the scale of the position to be discriminated, and determine the CTO detection model corresponding to this scale from the CTO detection model set, that is, determine the CTO detection model corresponding to the scale interval where this scale is located from the CTO detection model set. Thereafter, segment the associated image of the position to be discriminated from the coronary CTA image based on the two end points of the position to be discriminated, and input the associated image into the determined CTO detection model. It can be understood that the coronary CTA image contains rich data details. Therefore, it is necessary to expand in the coronary CTA image based on the two end points of the position to be discriminated to obtain an image with a relatively high correlation with the position to be discriminated (i.e., the associated image) and input it into the CTO detection model to obtain a more accurate detection result.

[0042] In specific applications, the associated image of the position to be discriminated can be obtained in the following way: In the coronary CTA image, expand respectively from the two end points of the position to be discriminated in the direction away from the position to be discriminated (i.e., expand to the outside of the position to be discriminated) until the three-dimensional region included between the two expanded parts reaches a preset first three-dimensional size. Among them, the first three-dimensional size is a relatively large size. For example, it is 64*64*64. Finally, determine the image of the three-dimensional region that reaches the first three-dimensional size as the associated image of the position to be discriminated.

[0043] After inputting the above-mentioned associated image into the corresponding CTO detection model, if the CTO detection model outputs a positive result, that is, outputs "CTO occurs", it is determined that CTO has occurred at the position to be discriminated. It can be understood that CTOs of different scales often correspond to different overall vascular connection situations. The present invention uses a CTO detection model matching the CTO scale to detect the position to be discriminated, which helps to improve the accuracy of CTO detection.

[0044] In practical applications, when the aforementioned vascular segmentation model extracts coronary blood vessels in the coronary CTA image, there is a certain possibility of overestimating the scale of the position to be discriminated. For example, the actual bloodless length of a certain position to be discriminated is 8 millimeters, and after being processed by the vascular segmentation model, it outputs 12 millimeters. This will affect the selection of the CTO detection model and further affect the accuracy of CTO detection. To solve the above problems, the present invention adopts the following method.

[0045] Specifically, if the foregoing CTO detection model outputs a negative result (i.e., it is determined that there is no CTO), the two endpoints of the position to be discriminated are respectively contracted by a preset distance along the direction close to the center of the position to be discriminated (i.e., contracted inward to the position to be discriminated), to obtain two new endpoints of the position to be discriminated. The above preset strategy can be flexibly set according to actual needs. For example, it can be set that the two endpoints are respectively contracted inward by 1 / 6 of the original scale of the position to be discriminated. In this way, if the original scale of the position to be discriminated is 12 millimeters, the scale after contraction is 8 millimeters. Figure 3 is a schematic diagram of the contraction of the position to be discriminated in an embodiment of the present invention. In Figure 3 it, the blood vessel before contraction is on the left, with endpoints A and B, and the blood vessel after contraction is on the right, with new endpoints C and D.

[0046] After obtaining the two new endpoints of the position to be discriminated, the scale between the two new endpoints can be calculated, and the CTO detection model corresponding to this scale can be queried from the CTO detection model set. Finally, the corresponding associated image is segmented from the coronary CTA image based on the two new endpoints, the associated image is input into the queried CTO detection model, and the output result of the CTO detection model is used as the CTO discrimination result of the position to be discriminated. That is, if the CTO detection model outputs "there is a CTO", it is determined that there is a CTO at the position to be discriminated; if the CTO detection model outputs "there is no CTO", it is determined that there is no CTO at the position to be discriminated. In this way, determining the corresponding CTO detection model according to the scale after contraction and performing re-detection can avoid missed detection of CTOs and help obtain more accurate CTO detection results.

[0047] In one embodiment, the above associated image based on the two new endpoints can be obtained through the following steps: First, in the coronary CTA image, expand respectively from the two new endpoints along the direction away from the position to be discriminated until the three-dimensional region included between the two expanded parts reaches a preset second three-dimensional size. The second three-dimensional size is a relatively small size, and is smaller than the first three-dimensional size in each dimension of length, width, and height. For example, the second three-dimensional size is 32*32*32. Thereafter, the image of the three-dimensional region that reaches the second three-dimensional size is determined as the associated image corresponding to the two new endpoints. By using the above expansion method and the above three-dimensional size setting method, it is possible to reduce the data operation amount on the basis of retaining sufficient image details, so as to achieve the balance of CTO detection accuracy and detection efficiency.

[0048] Figure 4 is a schematic diagram of the specific execution steps of the coronary CTA processing method for vascular occlusion lesions in an embodiment of the present invention. The steps shown therein have been described above and will not be repeated here.

[0049] In the technical solution of the embodiment of the present invention, first, the endpoints of the coronary artery vessels are identified from the coronary CTA images; then, the position to be discriminated suspected of having chronic total occlusion (CTO) is identified, and the position to be discriminated is the blood vessel without blood flow signal between two adjacent endpoints; finally, the position to be discriminated is detected by using a CTO detection model set including multiple CTO detection models corresponding to different CTO scale intervals, and the CTO discrimination result of the position to be discriminated is obtained. Through the above steps, it is possible to perform targeted judgment using the CTO detection model adapted to the scale of the position to be discriminated in the set, thereby improving the accuracy of CTO detection. In addition, in the existing blood vessel segmentation process, the scale of the position to be discriminated is easily enlarged, resulting in the subsequent use of a CTO detection model with a mismatched scale. To address the above defect, after detecting based on the original scale of the position to be discriminated, the present invention shrinks the position to be discriminated according to a preset strategy, determines the corresponding CTO detection model according to the shrunk scale, and performs re-detection, thereby avoiding missed detection of CTO and helping to obtain a more accurate CTO detection result.

[0050] It should be noted that for the foregoing method embodiments, for the sake of description, they are expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence. In fact, some steps can be performed in other sequences or simultaneously. In addition, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential for implementing the present invention.

[0051] To facilitate better implementation of the above solution of the embodiment of the present invention, the following also provides a related device for implementing the above solution.

[0052] Please refer to Figure 5 As shown, the coronary CTA processing device 500 for vascular occlusion lesions provided by the embodiment of the present invention may include: a vascular endpoint extraction unit 501, a position to be discriminated acquisition unit 502, and a CTO discrimination unit 503.

[0053] Among them, the vascular endpoint extraction unit 501 is used to identify the endpoints of the coronary artery vessels from the coronary CTA images; the position to be discriminated acquisition unit 502 is used to: identify the position to be discriminated suspected of having chronic total occlusion (CTO); where the position to be discriminated is the blood vessel without blood flow signal between two adjacent endpoints; the CTO discrimination unit 503 is used to: detect the position to be discriminated by using a pre-determined CTO detection model set, and obtain the CTO discrimination result of the position to be discriminated; where the CTO detection model set includes multiple CTO detection models pre-trained corresponding to different CTO scale intervals.

[0054] In an embodiment of the present invention, the vascular endpoint extraction unit 501 may further be configured to: input the coronary CTA image into a pre-trained vascular segmentation model to obtain the coronary blood vessels in the coronary CTA image; perform skeleton extraction on the coronary blood vessels, and determine the centerline of the coronary blood vessels according to the skeleton extraction result; wherein, the pixel connection of the centerline is a tree structure, and the root node of the tree structure is a pixel of the aorta; traverse each pixel in the tree structure from the root node, and determine the non-root node pixel with only one adjacent pixel as the endpoint of the coronary blood vessels.

[0055] In a specific application, the device may further include a training unit, which is configured to train any CTO detection model through the following steps: obtain CTO lesion images and non-CTO images; wherein, the CTO scale in the CTO lesion images is within a preset CTO scale range for this CTO detection model; use the CTO lesion images and the non-CTO images as training samples, and whether the CTO lesion images or the non-CTO images have CTO as labels to train this CTO detection model.

[0056] In an actual application, the CTO discrimination unit 503 may further be configured to: calculate the scale of the position to be discriminated, and determine the CTO detection model corresponding to this scale from the CTO detection model set; segment the associated image of the position to be discriminated from the coronary CTA image based on the two endpoints of the position to be discriminated, and input the associated image into the determined CTO detection model; if the CTO detection model outputs an affirmative result, then use this result as the CTO discrimination result of the position to be discriminated.

[0057] As a preferred solution, the coronary CTA image is a three-dimensional image; the CTO discrimination unit 503 may further be configured to: in the coronary CTA image, respectively expand from the two endpoints of the position to be discriminated in a direction away from the position to be discriminated until the three-dimensional region included between the two expanded parts reaches a preset first three-dimensional size; determine the image of the three-dimensional region that reaches the first three-dimensional size as the associated image of the position to be discriminated.

[0058] Preferably, the CTO discrimination unit 503 may further be configured to: if the determined CTO detection model outputs a negative result, then respectively contract the two endpoints of the position to be discriminated by a preset distance in a direction close to the center of the position to be discriminated to obtain two new endpoints of the position to be discriminated; calculate the scale between the two new endpoints, and query the CTO detection model corresponding to this scale from the CTO detection model set; segment the corresponding associated image from the coronary CTA image based on the two new endpoints, input the associated image into the queried CTO detection model, and use the output result of this CTO detection model as the CTO discrimination result of the position to be discriminated.

[0059] In addition, in the embodiment of the present invention, the CTO discrimination unit 503 can be further configured to: in the coronary CTA image, respectively expand from the two new endpoints in a direction away from the position to be discriminated until the three-dimensional region included between the two expanded parts reaches a preset second three-dimensional size; wherein, the second three-dimensional size is smaller than the first three-dimensional size in each dimension; and determine the image of the three-dimensional region that reaches the second three-dimensional size as the associated image corresponding to the two new endpoints.

[0060] In the technical solution of the embodiment of the present invention, first, the endpoints of the coronary blood vessels are identified from the coronary CTA image; then, the position to be discriminated where chronic total occlusion (CTO) is suspected to occur is identified, and the position to be discriminated is the blood vessel without blood flow signal between two adjacent endpoints; finally, the position to be discriminated is detected by using a CTO detection model set including a plurality of CTO detection models corresponding to different CTO scale intervals, and a CTO discrimination result of the position to be discriminated is obtained. Through the above steps, it is possible to use the CTO detection model adapted to the scale of the position to be discriminated in the set for targeted judgment, thereby improving the accuracy of CTO detection. In addition, in the existing blood vessel segmentation process, the scale of the position to be discriminated is easily enlarged, resulting in the subsequent use of a CTO detection model with an inconsistent scale. To address the above defects, after detecting based on the original scale of the position to be discriminated, the present invention shrinks the position to be discriminated according to a preset strategy, determines the corresponding CTO detection model according to the shrunk scale, and performs re-detection, thereby avoiding missed detection of CTO and helping to obtain a more accurate CTO detection result.

[0061] Figure 6 An exemplary system architecture 600 to which the coronary CTA processing method for vascular occlusion lesions or the coronary CTA processing device for vascular occlusion lesions according to the embodiment of the present invention can be applied is shown.

[0062] As Figure 6 shown, the system architecture 600 may include terminal devices 601, 602, 603, a network 604, and a server 605 (this architecture is only an example, and the components included in the specific architecture can be adjusted according to the specific situation of the application). The network 604 is used to provide a medium for communication links between the terminal devices 601, 602, 603 and the server 605. The network 604 may include various connection types, such as wired, wireless communication links, or fiber optic cables, etc.

[0063] Users can use the terminal devices 601, 602, 603 to interact with the server 605 through the network 604 to receive or send messages, etc. Various client applications, such as a CTO detection application (only an example), may be installed on the terminal devices 601, 602, 603.

[0064] The terminal devices 601, 602, and 603 can be various electronic devices with a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and so on.

[0065] The server 605 can be a server that provides various services, such as a background server (only an example) that supports the CTO detection application operated by the user using the terminal devices 601, 602, and 603. The background server can process the received CTO detection request and feedback the processing result (such as whether there is a CTO - only an example) to the terminal devices 601, 602, and 603.

[0066] It should be noted that the coronary CTA processing method for vascular occlusion lesions provided by the embodiments of the present invention is generally executed by the server 605. Correspondingly, the coronary CTA processing device for vascular occlusion lesions is generally provided in the server 605.

[0067] It should be understood that Figure 6 the numbers of the terminal devices, network, and server in

[0068] The present invention also provides an electronic device. The electronic device according to the embodiments of the present invention includes: one or more processors; a storage device for storing one or more programs, and when the one or more programs are executed by the one or more processors, the one or more processors implement the coronary CTA processing method for vascular occlusion lesions provided by the present invention.

[0069] Next, refer to Figure 7 , which shows a schematic structural diagram of a computer system 700 suitable for implementing the electronic device of the embodiments of the present invention. Figure 7 The shown electronic device is only an example and should not bring any limitation to the functions and usage scope of the embodiments of the present invention.

[0070] As Figure 7 shown, the computer system 700 includes a central processing unit (CPU) 701, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM) 702 or the program loaded from the storage section 708 into the random access memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the computer system 700 are also stored. The CPU 701, ROM 702, and RAM 703 are connected to each other through a bus 704. The input / output (I / O) interface 705 is also connected to the bus 704.

[0071] The following components are connected to the I / O interface 705: an input section 706 including a keyboard, a mouse, etc.; an output section 707 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc. and a speaker, etc.; a storage section 708 including a hard disk, etc.; and a communication section 709 including a network interface card such as a LAN card, a modem, etc. The communication section 709 performs communication processing via a network such as the Internet. A drive 710 is also connected to the I / O interface 705 as required. A removable medium 711 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is installed on the drive 710 as required so that a computer program read therefrom is installed into the storage section 708 as required.

[0072] Specifically, according to the embodiments disclosed in the present invention, the process described in the above main step diagram can be implemented as a computer software program. For example, the embodiments of the present invention include a computer program product which includes a computer program carried on a computer-readable medium, and the computer program includes program codes for performing the method shown in the main step diagram. In the above embodiments, the computer program can be downloaded and installed from the network through the communication section 709, and / or installed from the removable medium 711. When the computer program is executed by the central processing unit 701, the above functions defined in the system of the present invention are executed.

[0073] It should be noted that the computer-readable medium shown in the present invention can be a computer-readable signal medium, a computer-readable storage medium, or any combination of the above two. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, 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 disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in conjunction with an instruction execution system, apparatus, or device. In the present invention, the computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium can also be any computer-readable medium other than the computer-readable storage medium, and this computer-readable medium can send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wireless, wire, optical cable, RF, etc., or any suitable combination of the above.

[0074] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram can represent a module, a program segment, or a part of code, and the above module, program segment, or part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks can occur in a different order from that marked in the accompanying drawings. For example, two consecutive blocks shown can actually be executed substantially in parallel, and they can sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram or flowchart, and the combination of blocks in the block diagram or flowchart, can be implemented by a dedicated hardware-based system for performing the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0075] The units involved in the embodiments of the present invention can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes a coronary artery endpoint extraction unit, a position to be determined acquisition unit, and a CTO determination unit. Among them, the names of these units do not constitute a limitation to the unit itself in some cases. For example, the coronary artery endpoint extraction unit can also be described as "a unit that provides coronary artery endpoints to the position to be determined acquisition unit".

[0076] As another aspect, the present invention also provides a computer-readable medium. The computer-readable medium can be included in the device described in the above embodiments; it can also exist alone without being assembled into the device. The above computer-readable medium carries one or more programs. When the above one or more programs are executed by the device, the steps performed by the device include: identifying the endpoints of the coronary artery in the coronary CTA image; identifying the position to be determined where chronic total occlusion (CTO) is suspected to occur, where the position to be determined is a blood vessel without blood flow signal between two adjacent endpoints; using a pre-determined set of CTO detection models to detect the position to be determined and obtaining the CTO determination result of the position to be determined, where the set of CTO detection models includes multiple pre-trained CTO detection models corresponding to different CTO scale intervals.

[0077] In the technical solution of the embodiments of the present invention, first, the endpoints of the coronary artery are identified from the coronary CTA image; then, the position to be determined where chronic total occlusion (CTO) is suspected to occur is identified, and the position to be determined is a blood vessel without blood flow signal between two adjacent endpoints; finally, the position to be determined is detected using a set of CTO detection models including multiple CTO detection models corresponding to different CTO scale intervals, and the CTO determination result of the position to be determined is obtained. Through the above steps, it is possible to use the CTO detection model adapted to the scale of the position to be determined in the set for targeted judgment, thereby improving the accuracy of CTO detection. In addition, the existing blood vessel segmentation process is prone to expanding the scale of the position to be determined, resulting in the subsequent use of a CTO detection model with an inconsistent scale. To address the above defects, after detecting based on the original scale of the position to be determined, the present invention shrinks the position to be determined according to a preset strategy, determines the corresponding CTO detection model according to the shrunk scale, and performs re-detection, thereby avoiding CTO missed detection and helping to obtain a more accurate CTO detection result.

[0078] 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 occur depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A coronary CTA processing method for vascular occlusion lesions, characterized in that, Including: Identifying the endpoints of coronary arteries from coronary CTA images; Identifying a position to be discriminated that is suspected of having a chronic total occlusion (CTO); wherein, the position to be discriminated is a blood vessel without blood flow signal between two adjacent endpoints; Detecting the position to be discriminated using a pre-determined set of CTO detection models to obtain a CTO discrimination result for the position to be discriminated, including: calculating the scale of the position to be discriminated, and determining a CTO detection model corresponding to the scale from the set of CTO detection models; segmenting an associated image of the position to be discriminated from the coronary CTA image based on the two endpoints of the position to be discriminated, and inputting the associated image into the determined CTO detection model; if the CTO detection model outputs a positive result, using this result as the CTO discrimination result for the position to be discriminated; Wherein, the set of CTO detection models includes multiple pre-trained CTO detection models corresponding to different CTO scale intervals, and the CTO scale refers to the CTO occlusion length in the blood vessel; the multiple CTO detection models are similar models, each containing a convolutional layer for extracting image features and a classification layer for classifying according to the extracted image features; different CTO detection models have convolutional layers with different parameters and the same classification layer; Any CTO detection model is trained through the following steps: obtaining CTO lesion images and non-CTO images; wherein, the CTO scale in the CTO lesion images is within the CTO scale interval preset for this CTO detection model; using the CTO lesion images and the non-CTO images as training samples, and whether the CTO lesion images or the non-CTO images have CTO as labels to train this CTO detection model.

2. The method according to claim 1, wherein The identifying the endpoints of coronary arteries from coronary CTA images includes: Inputting the coronary CTA image into a pre-trained blood vessel segmentation model to obtain the coronary arteries in the coronary CTA image; Performing skeleton extraction on the coronary arteries, and determining the centerline of the coronary arteries based on the skeleton extraction result; wherein, the pixel connection of the centerline is a tree structure, and the root node of the tree structure is a pixel of the aorta; Traversing each pixel in the tree structure from the root node, and determining a non-root node pixel with only one adjacent pixel as the endpoint of the coronary artery.

3. The method according to claim 1, characterized in that The coronary CTA image is a three-dimensional image; and, the segmenting an associated image of the position to be discriminated from the coronary CTA image based on the two endpoints of the position to be discriminated includes: In the coronary CTA image, respectively expanding from the two endpoints of the position to be discriminated in a direction away from the position to be discriminated until the three-dimensional region included between the two expanded parts reaches a preset first three-dimensional size; Determining the image of the three-dimensional region that reaches the first three-dimensional size as the associated image of the position to be discriminated.

4. The method according to claim 3, wherein The detecting the position to be discriminated using a pre-determined set of CTO detection models to obtain a CTO discrimination result for the position to be discriminated further includes: If the determined CTO detection model outputs a negative result, the two endpoints of the position to be discriminated are respectively contracted by a preset distance along the direction close to the center of the position to be discriminated, so as to obtain two new endpoints of the position to be discriminated; Calculate the scale between the two new endpoints, and query the CTO detection model corresponding to this scale from the CTO detection model set; Based on the two new endpoints, segment the corresponding associated image from the coronary CTA image, input the associated image into the queried CTO detection model, and use the output result of the CTO detection model as the CTO discrimination result of the position to be discriminated.

5. The method according to claim 4, wherein The segmenting the corresponding associated image from the coronary CTA image based on the two new endpoints includes: In the coronary CTA image, expand from the two new endpoints along the direction away from the position to be discriminated respectively until the three-dimensional region included between the two expanded parts reaches a preset second three-dimensional size; wherein, the second three-dimensional size is smaller than the first three-dimensional size in each dimension; Determine the image of the three-dimensional region reaching the second three-dimensional size as the associated image corresponding to the two new endpoints.

6. A coronary CTA processing device for vascular occlusion lesions, characterized in that, Includes: A vascular endpoint extraction unit, configured to identify the endpoints of the coronary blood vessels from the coronary CTA image; A position to be discriminated acquisition unit, configured to: identify the position to be discriminated where chronic total occlusion (CTO) is suspected to occur; wherein, the position to be discriminated is a blood vessel without blood flow signal between two adjacent endpoints; A CTO discrimination unit, configured to: detect the position to be discriminated by using a pre-determined CTO detection model set to obtain the CTO discrimination result of the position to be discriminated; wherein, the CTO detection model set includes a plurality of pre-trained CTO detection models corresponding to different CTO scale intervals; the CTO scale refers to the CTO occlusion length in the blood vessel; the plurality of CTO detection models are similar models, all of which contain a convolutional layer for extracting image features and a classification layer for classifying according to the extracted image features; different CTO detection models have convolutional layers with different parameters and the same classification layer; A training unit, configured to train any CTO detection model through the following steps: obtain CTO lesion images and non-CTO images; wherein, the CTO scale in the CTO lesion images is within the CTO scale interval preset for this CTO detection model; use the CTO lesion images and the non-CTO images as training samples, and whether the CTO lesion images or the non-CTO images have CTO as labels to train this CTO detection model; The CTO discrimination unit is further configured to: calculate the scale of the position to be discriminated, determine the CTO detection model corresponding to this scale from the CTO detection model set; segment the associated image of the position to be discriminated from the coronary CTA image based on the two endpoints of the position to be discriminated, and input the associated image into the determined CTO detection model; if this CTO detection model outputs an affirmative result, use this result as the CTO discrimination result of the position to be discriminated.

7. An electronic device, characterized in that, Includes: One or more processors; A storage device, configured to store one or more programs, When the one or more programs are executed by the one or more processors such that the one or more processors implement the method according to any one of claims 1-5.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by a processor, the method according to any one of claims 1-5 is implemented.

Images