Vascular image processing methods, devices, readable storage media, and electronic devices

Abstract

This disclosure relates to a vascular image processing method, apparatus, readable storage medium, and electronic device. The method includes: acquiring a CTA image of a blood vessel; obtaining a three-dimensional segmentation result of the blood vessel based on the CTA image; extracting the centerline of the blood vessel based on the three-dimensional segmentation result; generating a local coordinate system for each point on the centerline based on a Rotation Minimizing Frame (RMF); and obtaining a three-dimensional straightened image of the blood vessel based on the local coordinate system. The three-dimensional straightened image of the blood vessel generated based on the local coordinate system of the RMF can clearly and intuitively present the morphology of the blood vessel, including the main artery, branches, and surrounding tissues. Different angles can be rotated around the central axis of the three-dimensional straightened image to obtain a two-dimensional straightened image at the corresponding angle, thereby achieving 360-degree comprehensive qualitative and quantitative analysis of the blood vessel, facilitating doctors to more quickly and accurately locate lesions.

Description

Technical Field

[0001] This disclosure relates to the field of image processing, and more specifically, to a method, apparatus, readable storage medium, and electronic device for processing vascular images. Background Technology

[0002] Aortic dissection and abdominal aortic aneurysm pose a serious threat to human health, and computed tomography angiography (CTA) imaging equipment has become an important diagnostic tool for these diseases. To address the visualization challenges of the aorta, curved surface reconstruction (CPR) technology is commonly used in image post-processing, highlighting the urgent need for a vascular image processing method to visually represent vascular morphology. Summary of the Invention

[0003] The purpose of this disclosure is to provide a method, apparatus, electronic device, and readable storage medium for processing vascular images.

[0004] According to a first aspect of the present disclosure, a vascular image processing method is provided, comprising: acquiring a CTA image of a vascular vessel; obtaining a three-dimensional segmentation result of the vascular vessel based on the CTA image; extracting the centerline of the vascular vessel based on the three-dimensional segmentation result; generating a local coordinate system for each point on the centerline based on a rotation-minimized frame; and obtaining a three-dimensional straightened image of the vascular vessel corresponding to the vascular vessel based on the local coordinate system.

[0005] Optionally, obtaining the three-dimensional segmentation result of the blood vessel based on the CTA image includes: obtaining a mask image of the blood vessel from the CTA image using a convolutional neural network; and superimposing the mask image to obtain the three-dimensional segmentation result of the blood vessel.

[0006] Optionally, the blood vessel is the aorta, which includes the ascending aorta and the descending aorta. Extracting the centerline of the blood vessel based on the three-dimensional segmentation result includes: obtaining the skeleton line of the aorta using a thinning algorithm; acquiring multiple endpoints on the skeleton line and generating an endpoint sequence; obtaining the starting point of the ascending aorta and the ending point of the descending aorta based on the endpoint sequence; extracting the initial centerline of the aorta based on the starting point using a shortest path method; and performing smoothing processing on the initial centerline after acquiring control points and spline interpolation to obtain the centerline.

[0007] Optionally, obtaining multiple endpoints on the skeleton line and generating an endpoint sequence includes: classifying each point on the skeleton line into endpoints, ordinary connection points, or bifurcation points based on the number of neighboring points of each point; and sorting the multiple endpoints on the skeleton line according to their positions in a specified direction to obtain an endpoint sequence.

[0008] Optionally, obtaining the starting point of the ascending aorta and the ending point of the descending aorta based on the endpoint sequence includes: taking the uppermost endpoint of the endpoint sequence with two connected regions as the starting point of the ascending aorta; and taking the lowermost endpoint of the endpoint sequence located in the descending aorta as the ending point.

[0009] Optionally, the method of extracting the initial centerline of the aorta based on the starting point using the shortest path includes: traversing the endpoints in the endpoint sequence sequentially starting from the starting point; stopping the traversal when the current endpoint is the termination point; deleting the bifurcation point when the current endpoint is a bifurcation point, and continuing to traverse the endpoints in the endpoint sequence until the termination point is reached; and extracting the initial centerline of the aorta based on the traversed endpoints.

[0010] Optionally, obtaining the three-dimensional straightened blood vessel image corresponding to the blood vessel according to the local coordinate system includes: acquiring the length and width of the output image specified by the user; obtaining the pixel values ​​of each two-dimensional cross-section of the blood vessel according to the local coordinate system; and superimposing each layer of the two-dimensional cross-section in the order of the points of the center line to obtain the three-dimensional straightened blood vessel image.

[0011] Optionally, after obtaining the three-dimensional straightened image of the blood vessel according to the local coordinate system, the process includes: rotating the three-dimensional straightened image of the blood vessel by different angles to obtain two-dimensional straightened images of the blood vessel at different angles.

[0012] According to a second aspect of the present disclosure, a vascular image processing apparatus is provided, comprising: an acquisition module for acquiring a CTA image of a vascular vessel; a segmentation module for obtaining a three-dimensional segmentation result of the vascular vessel based on the CTA image; a processing module for extracting the centerline of the vascular vessel based on the three-dimensional segmentation result; a generation module for generating a local coordinate system based on a rotation-minimized frame for each point on the centerline; the processing module is further configured to obtain a three-dimensional straightened vascular image corresponding to the vascular vessel based on the local coordinate system.

[0013] According to a third aspect of the present disclosure, a non-transitory computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the aforementioned vascular image processing method.

[0014] According to a fourth aspect of the present disclosure, an electronic device is provided, comprising: a memory having a computer program stored thereon; and a processor for executing the computer program in the memory to implement the steps of the aforementioned vascular image processing method.

[0015] The technical solutions provided by the embodiments of this disclosure can include the following beneficial effects: This disclosure can acquire CTA images of blood vessels; obtain three-dimensional segmentation results of blood vessels based on CTA images; extract the centerline of blood vessels based on the three-dimensional segmentation results; generate a local coordinate system for each point on the centerline based on a Rotation Minimizing Frame (RMF); and obtain a three-dimensional straightened image of the blood vessel based on the local coordinate system. The three-dimensional straightened image of blood vessels generated based on the local coordinate system of the RMF can clearly and intuitively present the morphology of blood vessels. The main artery, branches, and surrounding tissues of the blood vessels can be rotated around the central axis of the three-dimensional straightened image at different angles to obtain two-dimensional straightened images at the corresponding angles, thereby realizing 360-degree all-round qualitative and quantitative analysis of blood vessels, which makes it easier for doctors to discover the location of lesions more quickly and accurately.

[0016] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0017] The accompanying drawings are provided to further illustrate the present disclosure and form part of the specification. They are used together with the following detailed description to explain the present disclosure, but do not constitute a limitation thereof. In the drawings:

[0018] Figure 1 This is a schematic diagram of the structure of a computer system illustrated in an exemplary embodiment of this disclosure.

[0019] Figure 2 This is a flowchart illustrating an exemplary embodiment of the present disclosure of a blood vessel image processing method.

[0020] Figure 3 This is a flowchart illustrating another vascular image processing method according to an exemplary embodiment of this disclosure.

[0021] Figure 4 This is a schematic diagram of an aortic vessel as illustrated in an exemplary embodiment of this disclosure.

[0022] Figure 5 This is a schematic diagram of an axial cross-sectional image of an aorta, illustrating an exemplary embodiment of this disclosure.

[0023] Figure 6 This is a schematic diagram of a two-dimensional straightened image of an aorta, illustrating an exemplary embodiment of this disclosure.

[0024] Figure 7 This is a block diagram of a blood vessel image processing apparatus shown in an exemplary embodiment of the present disclosure.

[0025] Figure 8 This is a block diagram illustrating an electronic device according to an exemplary embodiment of the present disclosure. Detailed Implementation

[0026] The specific embodiments of this disclosure will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit this disclosure.

[0027] In the following description, when referring to the accompanying drawings, the same numbers in different drawings denote the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0028] Figure 1 This is a schematic diagram of the structure of a computer system shown in an exemplary embodiment of the present disclosure. The computer system includes a terminal 120 and a server 140.

[0029] Terminal 120 and server 140 are connected to each other via wired or wireless network.

[0030] Terminal 120 may include at least one of smartphones, laptops, desktop computers, tablets, smart speakers, and smart robots.

[0031] Terminal 120 includes a display; the display is used to display a three-dimensional image of the straightened blood vessel.

[0032] Terminal 120 includes a first memory and a first processor. The first memory stores a first program; the first program is invoked and executed by the first processor to implement the vascular image processing method provided in this disclosure. The first memory may include, but is not limited to, the following: Random Access Memory (RAM), Read-Only Memory (ROM), Programmable Read-Only Memory (PROM), Erasable Programmable Read-Only Memory (EPROM), and Electrically Erasable Programmable Read-Only Memory (EEPROM).

[0033] The first processor can consist of one or more integrated circuit chips. Optionally, the first processor can be a general-purpose processor, such as a central processing unit (CPU) or a network processor (NP).

[0034] Server 140 includes a second memory and a second processor. The second memory stores a second program, which is called by the second processor to implement the vascular image processing method provided in this disclosure. Optionally, the second memory may include, but is not limited to, the following: RAM, ROM, PROM, EPROM, and EEPROM. Optionally, the second processor may be a general-purpose processor, such as a CPU or NP.

[0035] The server can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN (Content Delivery Network), and big data and artificial intelligence platforms. The terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, etc., but is not limited to these. The terminal and server can be connected directly or indirectly through wired or wireless communication, and this disclosure does not impose any restrictions.

[0036] Over the past decade, endovascular aortic therapy has developed rapidly, achieving excellent therapeutic effects. The development of three-dimensional vessel straightening and reconstruction has played a crucial role in advancing this field. Most three-dimensional vessel straightening and reconstruction algorithms are based on the Frenet frame, consisting of a curve tangent vector, a principal normal vector, and a secondary normal vector. The secondary normal vector changes sign at inflection points, causing rotation of the principal and secondary normal vectors in the local coordinate systems of adjacent layers. This results in misalignment of adjacent layers in the output CTA image.

[0037] For the reasons stated above, this disclosure proposes a vascular image processing method to overcome the aforementioned problems. Please refer to [link / reference needed]. Figure 2 , Figure 2 This is a flowchart illustrating an exemplary embodiment of the present disclosure of a method for processing blood vessel images. The method is executed by a computer device, for example, by... Figure 1 The terminal or server in the computer system shown is used to execute the command. Figure 2 The illustrated vascular image processing method includes the following steps:

[0038] In step S101, CTA images of blood vessels are acquired.

[0039] CTA images refer to computed tomography angiography images. For example, CTA images of the aorta can be obtained.

[0040] In step S102, the three-dimensional segmentation result of the blood vessels is obtained based on the CTA image.

[0041] The contours of blood vessels in CTA images are extracted using relevant technical means as a 3D segmentation result. For example, the contours of blood vessels in CTA images can be extracted using neural networks.

[0042] In step S103, the centerline of the blood vessel is extracted based on the three-dimensional segmentation results.

[0043] The centerline is a line composed of the centers of multiple axial sections of the aorta. For example, the centerline of the blood vessel can be extracted from the three-dimensional segmentation results based on thinning algorithms and mathematical methods.

[0044] In step S104, a local coordinate system based on a rotation-minimized frame is generated for each point on the center line.

[0045] The Rotation Minimizing Frame (RMF) consists of three mutually perpendicular vectors. One vector is the tangent vector to the curve, and the other two form a plane perpendicular to the tangent vector (the normal plane). The normal plane is composed of a principal normal vector and a binormal vector. The principal normal vector points to the center of curvature at that point, while the binormal vector retains its sign at inflection points. The two normal vectors are consistent with the tangent vector of the curve. This minimizes the rotation of the plane around the centerline, preserving the continuity of the output image and preventing misalignment.

[0046] In step S105, a three-dimensional straightened image of the blood vessel is obtained according to the local coordinate system.

[0047] Since a local coordinate system was established along the centerline of the blood vessel, a series of normal planes of the centerline were obtained, which correspond to the axial cross-sectional images of the blood vessel. Superimposing these axial cross-sectional images yielded a cuboid around the centerline, which served as a three-dimensional straightened image of the blood vessel.

[0048] The three-dimensional straightened blood vessel image generated based on the local coordinate system of RMF can clearly and intuitively present the morphology of blood vessels, including the main arterial trunk, branches, and surrounding tissues. The image can be rotated around the central axis of the three-dimensional straightened blood vessel image at different angles to obtain the corresponding two-dimensional straightened image, thereby enabling qualitative and quantitative analysis of blood vessels from all angles, making it easier for doctors to quickly and accurately locate lesions.

[0049] Please see Figure 3 , Figure 3 This is a flowchart illustrating another vascular image processing method according to exemplary embodiments of this disclosure. The method is executed by a computer device, for example, by... Figure 1 The terminal or server in the computer system shown is used to execute the command.

[0050] It should be noted that, Figure 3 The vascular image processing method shown is similar to Figure 2 The implementation methods of the illustrated vascular image processing methods are consistent. Figure 3 For details not mentioned in the text, please refer to [the relevant source]. Figure 2 The description will not be repeated here. Figure 3 The illustrated vascular image processing method includes the following steps:

[0051] In step S201, CTA images of blood vessels are acquired.

[0052] For illustrative purposes, this disclosure uses the aorta as an example.

[0053] For example, raw aortic CTA images of aortic patients can be obtained using relevant medical tools, such as reading raw CTA images in .dcm format using the dicom tool and reading the grayscale values ​​of each layer of the raw CTA image as the aortic CTA image.

[0054] For example, the matrix corresponding to the grayscale values ​​can be represented as follows: Where a mn This represents the grayscale value in the m-th row and n-th column of a certain layer of the original CTA image.

[0055] In step S202, a mask image of the blood vessel is obtained from the CTA image using a convolutional neural network.

[0056] CTA images are preprocessed using a convolutional neural network to obtain the predicted segmentation result of the aorta in each layer of the aorta CTA image, which serves as the mask image for the aorta. The preprocessing process involves setting the grayscale value of the aorta region in each layer of the aorta CTA image as a first mask, and setting the grayscale value of other regions as a second mask. In one implementation, the first mask is 1 and the second mask is 0. For example, the mask image of the aorta can be represented as follows: Among them, b mn This indicates whether the pixel in row m and column n is a mask for the aorta. If it is the aorta, it is 1; otherwise, it is 0.

[0057] In step S203, the mask image is superimposed to obtain the three-dimensional segmentation result of the blood vessel.

[0058] The aortic prediction segmentation result of each layer of the aortic CTA image is obtained by using a convolutional neural network, which is the mask image of the aorta. The three-dimensional segmentation result of the aorta is obtained by superimposing each layer of the mask image.

[0059] In step S204, the centerline of the blood vessel is extracted based on the three-dimensional segmentation results.

[0060] The centerline of the aorta is extracted based on the three-dimensional segmentation results. The centerline is a line composed of the centers of multiple axial sections of the aorta.

[0061] Extracting the centerline of the aorta based on the 3D segmentation results involves the following five steps:

[0062] Step 1: Use a thinning algorithm to obtain the skeleton line of the aorta.

[0063] Thinning algorithms are used to thin contours in an image; that is, to extract a skeleton as quickly as possible while preserving the original image's topological structure. In this case, the skeleton line of the aorta is obtained through a thinning algorithm, facilitating subsequent processing of points along the skeleton line.

[0064] Step 2: Obtain multiple endpoints on the skeleton line and generate an endpoint sequence.

[0065] The points on the skeleton line are classified based on the number of neighboring points of each point. Each point is divided into endpoints, ordinary connection points, or branching points. For example, for a point, if the point has only 1 neighboring point, it is defined as an endpoint; if the point has 2 neighboring points, it is defined as an ordinary connection point; and if the point has 3 or more neighboring points, it is defined as a branching point.

[0066] After obtaining the endpoints, ordinary connection points, or bifurcation points on the skeleton line, the multiple endpoints on the skeleton line are sorted according to their positions in a specified direction to obtain an endpoint sequence. For example, each endpoint is read to obtain the i-th endpoint {E}. i =(x i y i , z i )|i=1,2,...,k},where, x i y i , z i Let be the values ​​of the i-th endpoint in the x-axis, y-axis, and z-axis directions, respectively, and k be the number of endpoints. Sort all endpoints according to their positions in the z-axis direction to obtain the endpoint sequence V = (E1, E2, ..., E...). k ).

[0067] Step 3: Obtain the starting point of the ascending aorta and the ending point of the descending aorta based on the endpoint sequence.

[0068] Based on the physiological structure of the aorta, it consists of a superior and a inferior portion. The superior portion includes the ascending and descending aorta. Therefore, the layer containing the endpoint of the ascending aorta comprises two connected regions. Please refer to [link to relevant documentation]. Figure 4 , Figure 4 This is a schematic diagram of an aortic vessel as illustrated in an exemplary embodiment of this disclosure.

[0069] from Figure 4 As can be seen, the starting point is located in the upper-middle part of the aorta. Therefore, based on the characteristics of the ascending aorta's physiological structure, the uppermost endpoint with two connected regions in the endpoint sequence can be taken as the starting point, and the lowermost endpoint in the descending aorta can be taken as the ending point. Because both the ascending and descending aortas are located on the same layer as the starting point, the starting point has two connected regions. For example, from the endpoint sequence on the ascending aorta, the maximum value Z of the endpoint sequence points in the z-axis direction can be obtained. max For E k The value along the z-axis, the minimum value Z min The value of E1 along the z-axis is used to determine the midpoint position as Z. mid =(Z max +Z min ) / 2, from Z mid To Z max Within the range, the endpoint layer has two connected regions, which are the starting points.

[0070] Similarly, based on the characteristics of the aorta's physiological structure, the lower half of the aorta, which includes the descending aorta, is a single vessel. Therefore, in the sequence of endpoints of the descending aorta, the lowest endpoint of the descending aorta is the termination point.

[0071] Step 4: Extract the initial centerline of the aorta based on the starting point using the shortest path method.

[0072] The method of extracting the initial centerline of the aorta based on the starting point and the shortest path includes: starting from the starting point, traversing the endpoints in the endpoint sequence in sequence. If the current endpoint is the termination point, stop traversing. If the current endpoint is the bifurcation point, delete the bifurcation point and continue traversing the endpoints in the endpoint sequence until the termination point is reached. The initial centerline of the aorta is extracted based on the traversed endpoints.

[0073] Step 5: After acquiring control points and spline interpolation for the initial centerline, perform smoothing processing to obtain the centerline.

[0074] Control points are acquired at equal intervals along the initial centerline of the aorta, and spline interpolation is performed. The initial centerline is then smoothed to obtain the final centerline. The intervals between the control points can be obtained based on empirical data or other feasible methods; this disclosure does not impose any restrictions on this. The number of spline interpolations can be, but is not limited to, three. Spline interpolation is a mathematical method for constructing a smooth curve passing through a series of known discrete points.

[0075] In step S205, a local coordinate system based on a rotation-minimized frame is generated for each point on the center line.

[0076] To eliminate the spatial curvature of the tubular aorta and describe tissue information near the aortic curve, it is necessary to establish the curvature of the imaging plane parallel to the centerline, i.e., to establish a local coordinate system along the aortic vessel centerline. In one implementation, the local coordinate system can be an RMF (Real-Time Factor Model).

[0077] The RMF coordinate system is primarily determined by the initial vector. In one implementation, for simplified calculation, the y-axis vector of the three-dimensional volume data (here, the three-dimensional coordinates corresponding to each point in the three-dimensional image of the aorta) can be selected as the reference vector to generate the initial vector. After obtaining the initial vector, the methods for calculating the tangent vector, principal normal vector, and binormal vector are as follows:

[0078] The centerline point set after collecting control points at equal intervals according to arc length is {C}. i =(x i y i , z i Let t be the number of points on the center line, where i = 1, 2, ..., N. The tangent vector t is obtained by the second-order difference of any center point in the center line point set. i =(C i+1 -C i-1 ) / 2, principal normal vector n i and the binormal vector b i The calculation formulas include:

[0079] n i =n i-1 -(n i-1 ·t i )t i

[0080] b i =t i *n i

[0081] Where · represents the dot product and * represents the cross product, the final result is the vector set {(t) in the RMF coordinate system. i n i b i )|i=1,2,...,N}.

[0082] In step S206, a three-dimensional straightened image of the blood vessel is obtained according to the local coordinate system.

[0083] Because the center of the RMF coordinate system is exactly at the center point of the centerline, the aorta is displayed in the center of the output image. Therefore, by specifying the length and width of the output image (generally, the length is set to equal the width), grayscale sampling can be performed using the principal and secondary normal vectors in the local coordinate system. Since a local coordinate system is established along the centerline, a series of normal planes of the centerline are obtained, corresponding to axial cross-sectional images of the aorta. Superimposing these axial cross-sectional images yields a cuboid surrounding the centerline, from which information about the aorta can be seen.

[0084] Using the RMF coordinate system obtained in the previous step, specify the length and width of the output image. For example, the length is equal to the width. The coordinates of the principal normal vector and the binormal vector of the i-th layer two-dimensional section (i.e., the axial section image) are x and x respectively. j and y j Where j = 1, 2, ..., M, and M is the length of the image, then within this two-dimensional cross-section (x j y j The corresponding 3D volume data coordinates at point ) are:

[0085] p j =C i +x j n i +y j b i

[0086] By using trilinear interpolation, the corresponding point p can be obtained. j The pixel value is obtained by trilinear interpolation, which is a method of linear interpolation on the tensor product grid of three-dimensional discrete sampled data.

[0087] Calculate the pixel value of each point within the two-dimensional section to obtain the pixel value of the i-th layer of the two-dimensional section, which corresponds to the axial cross-sectional image of the aorta, such as... Figure 5 As shown, Figure 5 This is a schematic diagram of an axial cross-sectional image of an aorta, illustrating an exemplary embodiment of this disclosure.

[0088] After obtaining the axial cross-sectional images of each point on the centerline of the aorta, N two-dimensional cross-sections are superimposed according to the order of the points on the centerline to obtain a surface reconstruction result, which is the three-dimensional straightened blood vessel image.

[0089] In step S207, the blood vessel is rotated by different angles around the central axis of the three-dimensional straightened image to obtain two-dimensional straightened images of the blood vessel at different angles.

[0090] The three-dimensional straightened blood vessel image generated in the previous step can be rotated around its central axis at different angles to obtain corresponding aortic vascular cross-sectional images, i.e., two-dimensional straightened images. This allows for 360-degree qualitative and quantitative analysis of the aorta. For example,... Figure 6 The image shown is a 2D straightened image at a 45-degree angle. Figure 6 This is a schematic diagram of a two-dimensional straightened image of an aorta, illustrating an exemplary embodiment of this disclosure.

[0091] It should be noted that this disclosure uses the aorta as an example for illustrative purposes, but the vascular image processing method shown in this disclosure can also be applied to other vascular images.

[0092] In summary, the vascular image processing method provided in this disclosure includes acquiring CTA images of blood vessels; obtaining three-dimensional segmentation results of blood vessels based on the CTA images; extracting the centerline of blood vessels based on the three-dimensional segmentation results; generating a local coordinate system for each point on the centerline based on a Rotation Minimizing Frame (RMF); and obtaining a three-dimensional straightened image of the blood vessel based on the local coordinate system. The three-dimensional straightened image of blood vessels generated based on the local coordinate system of the RMF can clearly and intuitively present the morphology of blood vessels. The main artery, branches, and surrounding tissues of the blood vessels can be rotated around the central axis of the three-dimensional straightened image at different angles to obtain two-dimensional straightened images at the corresponding angles, thereby achieving 360-degree comprehensive qualitative and quantitative analysis of blood vessels, facilitating doctors to more quickly and accurately locate lesions.

[0093] Figure 7 This is a block diagram of a blood vessel image processing apparatus illustrating an exemplary embodiment of this disclosure. (Refer to...) Figure 7 The vascular image processing device 20 includes an acquisition module 201, a segmentation module 202, a processing module 203, and a generation module 204.

[0094] The acquisition module 201 is used to acquire CTA images of blood vessels;

[0095] The segmentation module 202 is used to obtain the three-dimensional segmentation result of the blood vessel based on the CTA image;

[0096] The processing module 203 is used to extract the centerline of the blood vessel based on the three-dimensional segmentation result;

[0097] The generation module 204 is used to generate a local coordinate system for each point on the center line based on a rotation-minimized frame;

[0098] The processing module 203 is also used to obtain a three-dimensional straightened image of the blood vessel corresponding to the blood vessel based on the local coordinate system.

[0099] Optionally, the segmentation module 202 is further configured to obtain a mask image of the blood vessel from the CTA image using a convolutional neural network;

[0100] The three-dimensional segmentation result of the blood vessel is obtained by superimposing the mask image.

[0101] Optionally, the processing module 203 is further configured to obtain the skeleton line of the aorta using a thinning algorithm;

[0102] Obtain multiple endpoints on the skeleton line and generate an endpoint sequence;

[0103] The starting point of the ascending aorta and the ending point of the descending aorta are obtained based on the endpoint sequence.

[0104] Based on the starting point, the initial centerline of the aorta is extracted using the shortest path method;

[0105] After acquiring control points and performing spline interpolation on the initial centerline, a smoothing process is performed to obtain the centerline.

[0106] Optionally, the processing module 203 is further configured to classify each point into endpoints, ordinary connection points, or fork points based on the number of neighboring points of each point on the skeleton line.

[0107] The multiple endpoints on the skeleton line are sorted according to their positions in a specified direction to obtain an endpoint sequence.

[0108] Optionally, the processing module 203 is further configured to use the uppermost endpoint of the two connected regions in the endpoint sequence as the starting point of the ascending aorta;

[0109] The lowest endpoint in the descending aorta in the endpoint sequence is taken as the termination point.

[0110] Optionally, the processing module 203 is further configured to traverse the endpoints in the endpoint sequence sequentially starting from the starting point;

[0111] If the current endpoint is the termination point, stop traversing;

[0112] If the current endpoint is a branch point, delete the branch point and continue traversing the endpoints in the endpoint sequence until the termination point is reached.

[0113] The initial centerline of the aorta is extracted based on the endpoints traversed.

[0114] Optionally, the processing module 203 is also used to obtain the length and width of the output image specified by the user;

[0115] Based on the local coordinate system, the pixel values ​​of each two-dimensional section of the blood vessel are obtained;

[0116] The three-dimensional straightened blood vessel image is obtained by superimposing the two-dimensional cross-sections of each layer in the order of the points of the center line.

[0117] Optionally, the processing module 203 is also used to rotate the three-dimensional straightened blood vessel image by different angles to obtain two-dimensional straightened blood vessel images at different angles.

[0118] Regarding the apparatus in the above embodiments, the specific manner in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0119] Figure 8 This is a block diagram illustrating an electronic device 400 according to an exemplary embodiment. Figure 8 As shown, the electronic device 400 can be Figure 1 The terminal shown, the electronic device 400, may include: a processor 401, a memory 402. The electronic device 400 may also include one or more of a multimedia component 403, an input / output (I / O) interface 404, and a communication component 405.

[0120] The processor 401 controls the overall operation of the electronic device 400 to complete all or part of the steps in the aforementioned vascular image processing method. The memory 402 stores various types of data to support the operation of the electronic device 400. This data may include, for example, instructions for any application or method operating on the electronic device 400, and application-related data such as contact data, sent and received messages, images, audio, video, etc. The memory 402 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. Multimedia component 403 may include a screen and an audio component. The screen may be, for example, a touchscreen, and the audio component is used to output and / or input audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in memory 402 or transmitted via communication component 405. The audio component also includes at least one speaker for outputting audio signals. I / O interface 404 provides an interface between processor 401 and other interface modules, such as a keyboard, mouse, buttons, etc. These buttons may be virtual or physical buttons. Communication component 405 is used for wired or wireless communication between the electronic device 400 and other devices. Wireless communication, such as Wi-Fi, Bluetooth, Near Field Communication (NFC), 2G, 3G, 4G, NB-IoT, eMTC, or other 5G technologies, or combinations thereof, is not limited here. Therefore, the corresponding communication component 405 may include: a Wi-Fi module, a Bluetooth module, an NFC module, etc.

[0121] In an exemplary embodiment, the electronic device 400 may be implemented by one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), controllers, microcontrollers, microprocessors, or other electronic components to perform the above-described vascular image processing method.

[0122] In another exemplary embodiment, a computer-readable storage medium including program instructions is also provided, which, when executed by a processor, implement the steps of the aforementioned vascular image processing method. For example, the computer-readable storage medium may be the aforementioned memory 402 including program instructions, which may be executed by the processor 401 of the electronic device 400 to complete the aforementioned vascular image processing method.

[0123] In another exemplary embodiment, a computer program product is also provided, the computer program product comprising a computer program executable by a programmable device, the computer program having a code portion for performing the above-described vascular image processing method when executed by the programmable device.

[0124] The preferred embodiments of this disclosure have been described in detail above with reference to the accompanying drawings. However, this disclosure is not limited to the specific details of the above embodiments. Within the scope of the technical concept of this disclosure, various simple modifications can be made to the technical solutions of this disclosure, and these simple modifications all fall within the protection scope of this disclosure.

[0125] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, this disclosure will not describe the various possible combinations separately.

[0126] Furthermore, various different embodiments of this disclosure can be combined in any way, as long as they do not violate the spirit of this disclosure, they should also be regarded as the content disclosed in this disclosure.

Claims

1. A method for processing blood vessel images, characterized in that, include: Obtain CTA images of blood vessels; The three-dimensional segmentation result of the blood vessel is obtained based on the CTA image; The centerline of the blood vessel is extracted based on the three-dimensional segmentation results; A local coordinate system based on a rotation-minimized frame is generated for each point on the centerline. The rotation-minimized frame consists of three mutually perpendicular tangent vectors, a principal normal vector, and a secondary normal vector. The tangent vectors are the tangent vectors of the curve. The principal normal vector points to the center of curvature of the point on the centerline. The secondary normal vector does not change sign at inflection points. The principal normal vector, the secondary normal vector, and the tangent vector of the curve are kept consistent. The principal normal vector and the secondary normal vector form a normal plane perpendicular to the tangent vectors, and the rotation of the normal plane around the centerline is minimized. Based on the local coordinate system, a three-dimensional straightened image of the blood vessel is obtained; The blood vessel is the aorta, which includes the ascending aorta and the descending aorta. Extracting the centerline of the blood vessel based on the three-dimensional segmentation results includes: The skeleton line of the aorta is obtained using a thinning algorithm; Obtain multiple endpoints on the skeleton line and generate an endpoint sequence; The starting point of the ascending aorta and the ending point of the descending aorta are obtained based on the endpoint sequence; wherein, the starting point is the uppermost endpoint in the endpoint sequence that has two connected regions, and the ending point is the lowermost endpoint in the endpoint sequence that is located in the descending aorta. Based on the starting point, the initial centerline of the aorta is extracted using the shortest path method; After acquiring control points and performing spline interpolation on the initial centerline, a smoothing process is performed to obtain the centerline.

2. The method according to claim 1, characterized in that, The step of obtaining the three-dimensional segmentation result of the blood vessel based on the CTA image includes: The mask image of the blood vessel is obtained from the CTA image using a convolutional neural network; The three-dimensional segmentation result of the blood vessel is obtained by superimposing the mask image.

3. The method according to claim 1, characterized in that, The step of obtaining multiple endpoints on the skeleton line and generating an endpoint sequence includes: Based on the number of neighboring points of each point on the skeleton line, each point is divided into endpoints, ordinary connection points, or fork points; The multiple endpoints on the skeleton line are sorted according to their positions in a specified direction to obtain an endpoint sequence.

4. The method according to claim 1, characterized in that, The step of obtaining the starting point of the ascending aorta and the ending point of the descending aorta based on the endpoint sequence includes: The uppermost endpoint of the two connected regions in the endpoint sequence is taken as the starting point of the ascending aorta; The lowest endpoint in the descending aorta in the endpoint sequence is taken as the termination point.

5. The method according to claim 1, characterized in that, The method of extracting the initial centerline of the aorta based on the starting point using the shortest path includes: Starting from the starting point, traverse the endpoints in the endpoint sequence sequentially; If the current endpoint is the termination point, stop traversing; If the current endpoint is a branch point, delete the branch point and continue traversing the endpoints in the endpoint sequence until the termination point is reached. The initial centerline of the aorta is extracted based on the endpoints traversed.

6. The method according to claim 1, characterized in that, The step of obtaining the three-dimensional straightened image of the blood vessel based on the local coordinate system includes: Get the length and width of the output image specified by the user; Based on the local coordinate system, the pixel values ​​of each two-dimensional section of the blood vessel are obtained; The three-dimensional straightened blood vessel image is obtained by superimposing the two-dimensional cross-sections of each layer in the order of the points of the center line.

7. The method according to claim 1, characterized in that, After obtaining the three-dimensional straightened image of the blood vessel according to the local coordinate system, the process includes: By rotating the three-dimensional straightened blood vessel image around its central axis by different angles, two-dimensional straightened images of the blood vessel at different angles are obtained.

8. A vascular image processing device, characterized in that, include: The acquisition module is used to acquire CTA images of blood vessels; A segmentation module is used to obtain the three-dimensional segmentation result of the blood vessel based on the CTA image; The processing module is used to extract the centerline of the blood vessel based on the three-dimensional segmentation results; A generation module is used to generate a local coordinate system for each point on the centerline based on a rotation-minimized frame. The rotation-minimized frame consists of three mutually perpendicular tangent vectors, a principal normal vector, and a secondary normal vector. The tangent vectors are the tangent vectors of the curve. The principal normal vector points to the center of curvature of the point on the centerline. The secondary normal vector does not change sign at inflection points. The principal normal vector, the secondary normal vector, and the tangent vector of the curve are kept consistent. The principal normal vector and the secondary normal vector form a normal plane perpendicular to the tangent vectors, and the rotation of the normal plane around the centerline is minimized. The processing module is also used to obtain a three-dimensional straightened image of the blood vessel corresponding to the blood vessel based on the local coordinate system. The blood vessel is the aorta, which includes the ascending aorta and the descending aorta. The processing module is also used to obtain the skeleton line of the aorta using a thinning algorithm. Obtain multiple endpoints on the skeleton line and generate an endpoint sequence; The starting point of the ascending aorta and the ending point of the descending aorta are obtained based on the endpoint sequence; wherein, the starting point is the uppermost endpoint in the endpoint sequence that has two connected regions, and the ending point is the lowermost endpoint in the endpoint sequence that is located in the descending aorta. Based on the starting point, the initial centerline of the aorta is extracted using the shortest path method; After acquiring control points and performing spline interpolation on the initial centerline, a smoothing process is performed to obtain the centerline.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-7.

10. An electronic device, characterized in that, include: A memory on which computer programs are stored; A processor for executing the computer program in the memory to implement the steps of the method according to any one of claims 1-7.

Images