Medical image recognition method and device, computer equipment and storage medium

Abstract

The invention relates to a medical image recognition method and device, computer equipment and a storage medium. The method comprises the following steps: performing principal component analysis on an initial connected domain of a tubular object in a region of interest of a medical image to obtain a principal component analysis result corresponding to the initial connected domain, and determining a target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain, and determining attribute information of the tubular object according to a principal component analysis result corresponding to the target connected domain. By adopting the method, the embedded tubular object can be identified from the medical image.

Description

Technical Field

[0001] The present application relates to the field of medical technology, and in particular to a medical image recognition method, apparatus, computer equipment and storage medium. Background Art

[0002] In the medical field, there are application scenarios in which tubular objects are placed into the patient's body. For example, for patients with spontaneous cerebral hemorrhage (abbreviated as cerebral hematoma), when the patient's bleeding volume is large, a minimally invasive puncture and drainage technique is usually used to place a drainage tube in the brain to achieve drainage of the intracerebral hemorrhage.

[0003] Furthermore, doctors need to identify the tubular object from the medical image based on the patient's postoperative medical image to assist the doctor in postoperative evaluation or postoperative treatment. Therefore, how to identify the implanted tubular object from the medical image is a key research topic for people in this field. Summary of the invention

[0004] Based on this, it is necessary to provide a medical image recognition method, device, computer equipment and storage medium that can identify the inserted tubular object from the medical image in order to solve the above technical problems.

[0005] In a first aspect, the present application provides a medical image recognition method, the method comprising:

[0006] Performing principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain the principal component analysis result corresponding to the initial connected domain;

[0007] According to the principal component analysis result corresponding to the initial connected domain, the target connected domain is determined from the initial connected domain;

[0008] The attribute information of the tubular object is determined according to the principal component analysis results corresponding to the target connected domain.

[0009] In one embodiment, the attribute information includes the diameter of the tubular object; the attribute information is determined according to the principal component analysis result corresponding to the target connected domain, including:

[0010] Determine the first projection surface and the second projection surface according to the principal component analysis result corresponding to the target connected domain; the first projection surface is determined according to the first principal component and the second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to the first principal component and the third principal component in the principal component analysis result corresponding to the target connected domain;

[0011] Determine a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of the first principal component in the first projection surface, and determine a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface;

[0012] The diameter of the tubular object is determined according to the first diameter and the second diameter.

[0013] In one embodiment, determining a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of a first principal component in the first projection surface, and determining a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface, comprises:

[0014] For each first candidate coordinate value, determine a first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface, and determine the first tube diameter according to the first difference corresponding to each first candidate coordinate value; and

[0015] For each second candidate coordinate value, a second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface is determined, and the second tube diameter is determined according to the second difference corresponding to each second candidate coordinate value.

[0016] In one embodiment, determining the diameter of the tubular object according to the first diameter and the second diameter includes:

[0017] If the difference between the first tube diameter and the second tube diameter is greater than a preset difference, the major axis length of the tubular object is determined according to the maximum value of the first tube diameter and the second tube diameter, and the minor axis length of the tubular object is determined according to the minimum value of the first tube diameter and the second tube diameter;

[0018] If the difference between the first tube diameter and the second tube diameter is not greater than the preset difference, the diameter of the tubular object is determined according to an average value of the first tube diameter and the second tube diameter.

[0019] In one embodiment, the attribute information includes a center line of the tubular object; and determining the attribute information of the tubular object according to a principal component analysis result corresponding to the target connected domain includes:

[0020] Based on the first projection surface, determining a first center line corresponding to the first projection surface;

[0021] Determine an initial center line corresponding to the target connected domain based on the second projection surface and the first center line;

[0022] The centerline of the tubular object is determined based on the initial centerline.

[0023] In one embodiment, determining the centerline of the tubular object according to the initial centerline includes:

[0024] Extending the first end and / or the second end of the initial centerline in the region of interest to determine a first candidate centerline;

[0025] Determine a candidate search range according to the first candidate center line and the diameter of the tubular object;

[0026] Determine a candidate connected domain according to pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold; the first preset threshold is determined according to each pixel value corresponding to the target connected domain;

[0027] The center line of the tubular object is determined according to the initial center line and the second candidate center line in the candidate connected domain.

[0028] In one embodiment, determining the centerline of the tubular object according to the initial centerline and the second candidate centerline in the candidate connected domain includes:

[0029] Determine the distance between the initial center line and the adjacent center line; the adjacent center line is a second candidate center line adjacent to the initial center line;

[0030] The centerline of the tubular object is determined based on the distance, the initial centerline, and the adjacent centerlines.

[0031] In one embodiment, determining the centerline of the tubular object according to the distance, the initial centerline, and the adjacent centerlines includes:

[0032] If the distance is less than the preset distance, the initial center line and the adjacent center line are connected to update the initial center line;

[0033] Return to the step of determining the initial center line and the distance between adjacent center lines until the distance is not less than a preset distance, and use the initial center line when the distance is not less than the preset distance as the center line of the tubular object.

[0034] In one embodiment, determining the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain includes:

[0035] The initial connected domain that meets the preset conditions is used as the target connected domain;

[0036] Among them, the preset conditions include that the ratio of a first length corresponding to the first principal component to the second length corresponding to the second principal component is greater than a first multiple, and the ratio of the first length to the third length corresponding to the third principal component is greater than a second multiple, the first length is determined according to the coordinate range corresponding to the first principal component, the second length is determined according to the coordinate range corresponding to the second principal component, and the third length is determined according to the coordinate range corresponding to the third principal component.

[0037] In one embodiment, the method further comprises:

[0038] Access to medical images;

[0039] Segment the medical image to obtain the region of interest;

[0040] The region of interest is segmented by a threshold value according to a second preset threshold value to obtain an initial connected region.

[0041] In one embodiment, the method further comprises:

[0042] Model the tubular object according to the attribute information to obtain a reconstruction result of the tubular object;

[0043] Reconstruction results showing tubular objects in medical images.

[0044] In a second aspect, the present application also provides a medical image recognition device, the device comprising:

[0045] The first determination module is used to perform principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain a principal component analysis result corresponding to the initial connected domain;

[0046] A second determination module is used to determine a target connected domain from the initial connected domain according to a principal component analysis result corresponding to the initial connected domain;

[0047] The third determination module is used to determine the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain.

[0048] In a third aspect, the present application further provides a computer device, wherein the computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of any of the above methods when executing the computer program.

[0049] In a fourth aspect, the present application further provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the steps of any of the above methods when executed by a processor.

[0050] In a fifth aspect, the present application further provides a computer program product, wherein the computer program product comprises a computer program, and when the computer program is executed by a processor, the steps of any of the above methods are implemented.

[0051] The above-mentioned medical image recognition method, device, computer equipment and storage medium perform principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain the principal component analysis result corresponding to the initial connected domain, and determine the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain, and then determine the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain. Since the initial connected domain can be screened according to the principal component analysis result corresponding to the initial connected domain, a more accurate target connected domain of the tubular object can be determined from the initial connected domain. Furthermore, after the attribute information of the tubular object is determined according to the principal component analysis result of the target connected domain, the shape, position and other information of the tubular object can be determined, thereby realizing efficient and accurate recognition of the tubular object from the medical image. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 This is a diagram of the application environment of the medical image recognition method in the embodiment of the present application;

[0053] Figure 2 This is a schematic diagram of the process of the medical image recognition method in the embodiment of the present application;

[0054] Figure 3 This is a schematic diagram of an initial connected domain in an embodiment of the present application;

[0055] Figure 4 This is a schematic diagram of a process for determining attribute information in an embodiment of the present application;

[0056] Figure 5 This is a schematic diagram of obtaining a projection surface in an embodiment of the present application;

[0057] Figure 6 This is a schematic diagram of the effects of a first projection surface and a second projection surface in an embodiment of the present application;

[0058] Figure 7 This is a schematic diagram of a process for determining a first pipe diameter and a second pipe diameter in an embodiment of the present application;

[0059] Figure 8 This is a schematic diagram of a process for determining a pipe diameter in an embodiment of the present application;

[0060] Fig. 9 This is another schematic diagram of a process for determining attribute information in an embodiment of the present application;

[0061] Fig.10 A schematic diagram of a two-dimensional center line in an embodiment of the present application;

[0062] Fig.11 This is a schematic diagram of a process for determining a center line in an embodiment of the present application;

[0063] Fig.12 This is a schematic diagram of an extended process in an embodiment of the present application;

[0064] Fig.13 A schematic diagram of a candidate connected domain in an embodiment of the present application;

[0065] Fig.14 This is another schematic diagram of a process for determining a center line in an embodiment of the present application;

[0066] Fig.15 This is another schematic diagram of a process for determining a center line in an embodiment of the present application;

[0067] Fig.16 This is a schematic diagram of a process of obtaining an initial connected domain in an embodiment of the present application;

[0068] Fig.17 This is a schematic diagram of a process for displaying reconstruction results in an embodiment of the present application;

[0069] Fig.18 A schematic diagram of a reconstruction result in an embodiment of the present application;

[0070] Fig.19 This is a diagram illustrating the principle of a medical image recognition method in an embodiment of the present application;

[0071] Fig. 20 A schematic diagram of a process of a medical image recognition method in an embodiment of the present application;

[0072] Fig.21 4 is a structural block diagram of a medical image recognition device in an embodiment of the present application. DETAILED DESCRIPTION

[0073] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0074] Taking the tubular object as the drainage tube in a patient with cerebral hematoma as an example, when the patient has a large amount of bleeding, a minimally invasive puncture and drainage technique is usually used to place a drainage tube in the brain to drain the intracerebral hemorrhage. Within 12 hours after the operation, the doctor needs to use the patient's medical images to confirm the placement of the drainage tube and the drainage of the hematoma to assist the doctor in postoperative evaluation or postoperative treatment.

[0075] However, currently different drainage tubes have different materials and imaging characteristics. In addition, most drainage tubes are made of soft materials and will bend and deform to a certain extent after entering the patient's brain tissue. Furthermore, the hollow structure and porous front end of the drainage tube make it difficult for doctors to identify the drainage tube from medical images.

[0076] Based on this, it is necessary to provide a medical image recognition method to address the above technical issues. The medical image recognition method will be introduced below.

[0077] Figure 1 FIG. 1 is an application environment diagram of the medical image recognition method in the embodiment of the present application. In the embodiment of the present application, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as shown in FIG. Figure 1As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store relevant data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a medical image recognition method is implemented.

[0078] Those skilled in the art will understand that Figure 1 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0079] This embodiment uses the method applied to a server as an example for illustration. It is understandable that the method can also be applied to a terminal, and can also be applied to a system including a terminal and a server, and is implemented through the interaction between the terminal and the server. The terminal can be, but is not limited to, various personal computers, laptops, smart phones, and tablet computers. The server can be implemented as an independent server or a server cluster consisting of multiple servers.

[0080] Figure 2 This is a flow chart of the medical image recognition method in the embodiment of the present application. The method can be applied to Figure 1 In the computer device shown, in one embodiment, as Figure 2 As shown, the following steps are included:

[0081] S201, performing principal component analysis on an initial connected domain of tubular objects in a region of interest of a medical image to obtain a principal component analysis result corresponding to the initial connected domain.

[0082] In this embodiment, the computer device can obtain the initial connected domain of the tubular object in the region of interest of the medical image. The medical image may include but is not limited to at least one of CT (Computed Tomography) images, MRI (Magnetic Resonance Imaging) images, PET (Positron Emission Computed Tomography)-CT images and PET-MR images. The medical image can be a two-dimensional image or a three-dimensional image. The region of interest refers to the area where the tubular object is inserted. The tubular object may include a hollow tubular structure. Exemplarily, the medical image is a CT image and the region of interest is brain tissue.

[0083] The initial connected domain refers to the complete area where the tubular object is located in the region of interest, which indicates the area where the tubular object may exist in the region of interest. It can be understood that the region of interest of the medical image may include multiple initial connected domains.

[0084] Optionally, the computer device can obtain a medical image sent by other devices, in which the initial connected domain of the tubular objects in the region of interest has been marked; the computer device can also obtain the medical image and then analyze the medical image to obtain the initial connected domain of the tubular objects in the region of interest of the medical image; the computer device can also determine the initial connected domain of the tubular objects in the region of interest from the medical image in response to the doctor's outlining or selection operation.

[0085] Figure 3 is a schematic diagram of an initial connected domain in an embodiment of the present application, such as Figure 3 As shown, the black areas are multiple initial connected domains of the tubular object in the region of interest 302 of the medical image 301 .

[0086] Afterwards, the computer device can perform principal component analysis (PCA) on the initial connected domain to obtain the principal component analysis results corresponding to the initial connected domain. The principal component analysis results corresponding to the initial connected domain can indicate the tubular shape of each initial connected domain, such as the length and width of the corresponding initial connected domain.

[0087] S202, determining a target connected domain from the initial connected domain according to a principal component analysis result corresponding to the initial connected domain.

[0088] The initial connected domain may not be accurate, and other tissues in the region of interest may also be considered as the initial connected domain of the tubular object. Therefore, in this embodiment, the computer device determines the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain. The target connected domain represents the area in the region of interest where the tubular object is most likely to exist.

[0089] Optionally, the computer device may determine a target connected domain in the initial connected domain whose length is greater than a preset length based on the principal component analysis result corresponding to the initial connected domain. In some embodiments, in order to further improve the accuracy, the computer device may also determine a target connected domain in the initial connected domain whose length is greater than a preset length and whose width is less than a preset width based on the principal component analysis result corresponding to the initial connected domain. This embodiment is not limited thereto.

[0090] Exemplarily, it is assumed that what is obtained in S201 is principal component analysis result 1 of initial connected domain 1, principal component analysis result 2 of initial connected domain 2, and principal component analysis result 3 of initial connected domain 3. Principal component analysis result 1 of initial connected domain 1 indicates that the length of initial connected domain 1 is A1 and the width is B1, principal component analysis result 2 of initial connected domain 1 indicates that the length of initial connected domain 2 is A2 and the width is B2, principal component analysis result 3 of initial connected domain 3 indicates that the length of initial connected domain 3 is A3 and the width is B3, and only if A1 is less than the preset length, the computer device can use initial connected domain 2 and initial connected domain 3 as target connected domains.

[0091] Please continue to refer to Figure 3 ,like Figure 3 As shown, after the target connected domain 303 is determined from the initial connected domain, the impurity region that does not conform to the structural characteristics of the tubular object is eliminated, and the region that conforms to the slender characteristics of the tubular object is selected.

[0092] S203, determining attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain.

[0093] Furthermore, the computer device can continue to determine the attribute information of the tubular object in the medical image according to the principal component analysis result corresponding to the target connected domain. The attribute information of the tubular object is used to describe at least one of the shape, position and size of the tubular object in the medical image.

[0094] Continuing with the above example, the target connected domain includes the initial connected domain 2 and the initial connected domain 3. The length A2 and the width B2 can be used as the attribute information of the tubular object corresponding to the initial connected domain 2, and the length A3 and the width B3 can be used as the attribute information of the tubular object corresponding to the initial connected domain 3.

[0095] In some embodiments, the computer device can mark the tubular object in the medical image according to the attribute information of the tubular object, so that the doctor can know the shape, position, size, etc. of the tubular object in the medical image.

[0096] The medical image recognition method provided in this embodiment performs principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain the principal component analysis result corresponding to the initial connected domain, and determines the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain, and then determines the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain. Since the initial connected domain can be screened according to the principal component analysis result corresponding to the initial connected domain, a more accurate target connected domain of the tubular object can be determined from the initial connected domain. Furthermore, after the attribute information of the tubular object is determined according to the principal component analysis result of the target connected domain, the shape, position and other information of the tubular object can be determined, thereby realizing efficient and accurate recognition of the tubular object from the medical image.

[0097] Figure 4 This is a schematic diagram of a process for determining attribute information in an embodiment of the present application, referring to Figure 4 This embodiment involves an optional implementation method of how to determine the attribute information. Based on the above embodiment, the attribute information includes the diameter of the tubular object; the above S203, determining the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain, includes the following steps:

[0098] S401, determining a first projection surface and a second projection surface according to the principal component analysis result corresponding to the target connected domain; the first projection surface is determined according to the first principal component and the second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to the first principal component and the third principal component in the principal component analysis result corresponding to the target connected domain.

[0099] In this embodiment, the principal component analysis results include analysis results in three directions, which are described with the three directions being the first principal component, the second principal component and the third principal component. The first principal component represents the extension direction of the tubular object, and the second principal component and the third principal component correspond to the two radial directions of the drainage tube, respectively.

[0100] Furthermore, the principal component analysis result may include the coordinate sets of the tubular object in the first principal component, the second principal component and the third principal component, respectively. In other words, the principal component analysis result includes the three-dimensional coordinate set of the region where the tubular object is located.

[0101] Furthermore, the computer device can determine the first projection surface and the second projection surface according to the principal component analysis result corresponding to the target connected domain. Figure 5 This is a schematic diagram of obtaining a projection surface in an embodiment of the present application, such as Figure 5As shown, taking the x direction representing the first principal component, the y direction representing the second principal component, and the z direction representing the third principal component as an example, cylinder 503 is a cylinder formed by the principal component analysis results of a target connected domain. By projecting the principal component analysis results on the plane formed by the first principal component and the second principal component, a first projection surface 501 can be obtained. By projecting the principal component analysis results on the plane formed by the first principal component and the second principal component, a first projection surface 502 can be obtained.

[0102] It is understandable that the first projection plane includes the two-dimensional coordinate set of the tubular object corresponding to the target connected domain on the plane formed by the x-direction and the y-direction. The second projection plane includes the two-dimensional coordinate set of the tubular object corresponding to the target connected domain on the plane formed by the x-direction and the z-direction. It should be noted that Figure 5 Taking the first projection surface and the second projection surface as rectangular shapes as an example, the actual first projection surface and the second projection surface are not necessarily standard rectangular shapes. Figure 6 This is a schematic diagram of the effects of a first projection surface and a second projection surface in an embodiment of the present application. According to different marks, Figure 6 A first projection surface 601 and a second projection surface 602 are shown.

[0103] S402, determining a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of the first principal component in the first projection surface, and determining a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface.

[0104] Furthermore, the computer device can select multiple first candidate coordinate values ​​on the first principal component based on the first projection surface. Figure 5 According to the first projection surface, the computer device can determine the coordinate value range on the first principal component. Furthermore, the computer device can traverse the coordinate value range on the first principal component according to a certain step length, or randomly select from the coordinate value range on the first principal component to obtain N first candidate coordinate values. Figure 5 X0 is a schematic representation of a first candidate coordinate value.

[0105] Similarly, for the second projection surface, the computer device selects M second candidate coordinate values ​​on the first principal component based on the second projection surface. It should be noted that the first candidate coordinate value and the second candidate coordinate value may be the same or different. N and M are both integers greater than 0, and N and M may be equal or unequal.

[0106] Furthermore, the computer device can determine the first tube diameter corresponding to the first projection surface based on multiple first candidate coordinate values ​​of the first principal component in the first projection surface, and determine the second tube diameter corresponding to the second projection surface based on multiple second candidate coordinate values ​​of the first principal component in the second projection surface.

[0107] Taking the first candidate coordinate value equal to X0 as an example, optionally, the computer device can draw a straight line along X0 and solve the line segment length of the straight line in the first projection plane, and the line segment length also represents the first candidate diameter of the target connected domain at X0. Similarly, N first candidate coordinate values ​​can correspond to N first candidate diameters. After that, the computer device can determine the first diameter according to the N first candidate diameters. For example, the computer device uses the average value of the N first candidate diameters as the first diameter.

[0108] The principle of the second pipe diameter is the same. M second candidate pipe diameters can determine M second candidate pipe diameters. Then, the computer device can determine the second pipe diameter according to the M second candidate pipe diameters.

[0109] S403, determining the diameter of the tubular object according to the first diameter and the second diameter.

[0110] Furthermore, the computer device can determine the diameter of the tubular object according to the first diameter and the second diameter. Optionally, the computer device can calculate the average or weighted average of the first diameter and the second diameter to obtain the diameter of the tubular object.

[0111] This embodiment determines the first projection surface and the second projection surface according to the principal component analysis result corresponding to the target connected domain. Since the first projection surface is determined according to the first principal component and the second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to the first principal component and the third principal component in the principal component analysis result corresponding to the target connected domain, the first diameter corresponding to the first projection surface can be determined according to the multiple first candidate coordinate values ​​of the first principal component in the first projection surface, and the second diameter corresponding to the second projection surface can be determined according to the multiple second candidate coordinate values ​​of the first principal component in the second projection surface, and finally the diameter of the tubular object can be determined according to the first diameter and the second diameter. Determining the attribute information of the tubular object on two orthogonal first and second projection surfaces reduces the data processing amount of the computer equipment and improves the recognition efficiency.

[0112] Figure 7 This is a schematic diagram of a process for determining the first pipe diameter and the second pipe diameter in an embodiment of the present application, referring to Figure 7 This embodiment relates to an optional implementation method of how to determine the first tube diameter and the second tube diameter. Based on the above embodiment, the above S402, according to the multiple first candidate coordinate values ​​of the first principal component in the first projection surface, determines the first tube diameter corresponding to the first projection surface, and according to the multiple second candidate coordinate values ​​of the first principal component in the second projection surface, determines the second tube diameter corresponding to the second projection surface, including the following steps:

[0113] S701, for each first candidate coordinate value, determine a first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface, and determine a first tube diameter according to the first difference corresponding to each first candidate coordinate value.

[0114] S702: for each second candidate coordinate value, determine a second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface, and determine the second tube diameter according to the second difference corresponding to each second candidate coordinate value.

[0115] In this embodiment, for N first candidate coordinate values, each first candidate coordinate value is substituted into the first projection surface, and the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface can be calculated. Figure 5 Take X0 in as an example, the minimum coordinate value corresponding to X0 is Figure 5 The maximum coordinate value corresponding to ymin, X0 shown in Figure 5 By subtracting ymax from ymin, a first difference between the minimum coordinate value and the maximum coordinate value of the second principal component can be determined. Similarly, N first candidate coordinate values ​​can determine N first differences.

[0116] Then, the computer device can determine the first pipe diameter according to the N first differences. For example, the computer device uses the average value, weighted average value or mode of the N first differences as the first pipe diameter.

[0117] Similarly, for the second pipe diameter, for the M second candidate coordinate values, each second candidate coordinate value is substituted into the second projection surface, and the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface can also be calculated. The second difference value can be obtained by subtracting the minimum coordinate value from the maximum coordinate value. The M second candidate coordinate values ​​can determine the M second difference values, and then the computer device can determine the second pipe diameter according to the M second difference values. For example, the computer device uses the average value, weighted average value or mode of the M second difference values ​​as the second pipe diameter.

[0118] In this embodiment, for each first candidate coordinate value, the first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface is determined, and the first pipe diameter is determined according to the first difference corresponding to each first candidate coordinate value. For each second candidate coordinate value, the second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface is determined, and the second pipe diameter is determined according to the second difference corresponding to each second candidate coordinate value. In this way, since multiple first candidate coordinate values ​​are selected in the first projection surface, the accuracy of the first pipe diameter can be improved. Similarly, multiple second candidate coordinate values ​​are selected in the second projection surface, which also improves the accuracy of the second pipe diameter.

[0119] Figure 8 This is a schematic diagram of a process for determining the pipe diameter in an embodiment of the present application, referring to Figure 8 This embodiment relates to an optional implementation method of how to determine the diameter of a tubular object. Based on the above embodiment, the above S403, determining the diameter of the tubular object according to the first diameter and the second diameter, includes the following steps:

[0120] S801, if the difference between the first tube diameter and the second tube diameter is greater than a preset difference, determining the major axis length of the tubular object according to the maximum value of the first tube diameter and the second tube diameter, and determining the minor axis length of the tubular object according to the minimum value of the first tube diameter and the second tube diameter.

[0121] After the above S701 and S702, after the computer device determines the first tube diameter corresponding to the first projection surface and the second tube diameter corresponding to the second projection surface, it further determines whether the difference between the first tube diameter and the second tube diameter is large.

[0122] The computer device determines the difference between the first pipe diameter and the second pipe diameter using the preset difference. The difference between the first pipe diameter and the second pipe diameter may include, but is not limited to, at least one of a difference, an absolute value of the difference, and a variance. The preset difference may be a value stored in advance in the computer device, or a value determined by the computer device based on the first pipe diameter or the second pipe diameter. For example, the computer device may use 10% of the first pipe diameter or 10% of the second pipe diameter as the preset difference.

[0123] Furthermore, if the difference between the first tube diameter and the second tube diameter is greater than the preset difference, it means that the shape of the target tubular object is greatly different on the first projection surface and the second projection surface, then the computer device considers that the tubular object corresponding to the target connected domain is an elliptical cylinder, and determines the major axis length of the tubular object according to the maximum value of the first tube diameter and the second tube diameter, and determines the minor axis length of the tubular object according to the minimum value of the first tube diameter and the second tube diameter.

[0124] For example, the computer device may use the maximum value of the first tube diameter and the second tube diameter as the long axis length of the tubular object, and use the minimum value of the second tube diameter and the second tube diameter as the short axis length of the tubular object. The computer device may also round up the maximum value of the first tube diameter and the second tube diameter as the long axis length of the tubular object, and round down the minimum value of the second tube diameter and the second tube diameter as the short axis length of the tubular object.

[0125] S802: If the difference between the first tube diameter and the second tube diameter is not greater than a preset difference, determine the diameter of the tubular object according to an average value of the first tube diameter and the second tube diameter.

[0126] If the difference between the first tube diameter and the second tube diameter is not greater than the preset difference, it means that the shape of the target tubular object is not much different on the first projection surface and the second projection surface, and the computer device considers that the tubular object corresponding to the target connected domain is a standard cylinder, and determines the diameter of the tubular object according to the average value of the first tube diameter and the second tube diameter. For example, the computer device can directly use the average value of the first tube diameter and the second tube diameter as the diameter of the tubular object, or round up or down the average value of the first tube diameter and the second tube diameter as the diameter of the tubular object.

[0127] It can be understood that if the tubular object corresponding to the target connected domain is determined to be a standard cylinder, the diameter of the tubular object is used to indicate the diameter of the standard cylinder; if the tubular object corresponding to the target connected domain is determined to be an elliptical cylinder, the diameter of the tubular object is used to indicate the major axis length and minor axis length of the elliptical cylinder.

[0128] In this embodiment, if the difference between the first tube diameter and the second tube diameter is greater than the preset difference, the long axis length of the tubular object is determined according to the maximum value of the first tube diameter and the second tube diameter, and the short axis length of the tubular object is determined according to the minimum value of the first tube diameter and the second tube diameter. If the difference between the first tube diameter and the second tube diameter is not greater than the preset difference, the diameter of the tubular object is determined according to the average value of the first tube diameter and the second tube diameter. Therefore, the shape of the tubular object can be flexibly adjusted according to the conditions of the first tube diameter and the second tube diameter, thereby improving the recognition accuracy of the tubular object.

[0129] Fig. 9 This is another flow chart of determining attribute information in an embodiment of the present application, referring to Fig. 9 This embodiment involves an optional implementation method of how to determine the attribute information of a tubular object. Based on the above embodiment, the attribute information includes the center line of the tubular object; the above S203, determining the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain, includes the following steps:

[0130] S901, based on the first projection surface, determining a first center line corresponding to the first projection surface.

[0131] In this embodiment, the center line of the tubular object can be determined using a quadratic smooth curve of the first projection surface and the second projection surface on the first principal component. The computer device can extract the first projection surface using the RANSAC (Random Sample Consensus) method to determine the first center line corresponding to the first projection surface. In some embodiments, the computer device can also determine the first center line corresponding to the first projection surface by other plane fitting methods.

[0132] S902: Determine an initial center line corresponding to the target connected domain based on the second projection surface and the first center line.

[0133] Afterwards, the computer device can substitute the first center line into the second projection surface, and continue to determine the initial center line corresponding to the target connected domain based on the second projection surface and the first center line. It can be understood that the initial center line is a three-dimensional center line.

[0134] Optionally, the computer device can use the RANSAC method to determine the initial center line based on the second projection surface and the first center line. Since the RANSAC method undergoes multiple random samplings, when the tubular object is bent, squeezed, or missing to a certain extent, the relatively complete target connected domain and the RANSAC method can be used to restore the tubular object relatively accurately, thereby improving the situation where some points in the tubular object are missing due to the openings, and improving the accuracy of the initial center line. In addition, the three-dimensional center line obtained by calculation based on the two-dimensional projection surface can reduce the computational complexity of the computer device.

[0135] It can be understood that the above example uses the first projection surface to first determine the first center line and then uses the second projection surface to determine the initial center line. In some embodiments, the second center line corresponding to the second projection surface can be first determined based on the second projection surface, and then the initial center line can be determined based on the first projection surface and the second center line. Fig.10 is a schematic diagram of a two-dimensional center line in an embodiment of the present application, such as Fig.10 As shown, in Figure 6 On the basis of, the first center line 1001 can be determined based on the first projection surface 601, and the second center line 1002 can be determined based on the second projection surface 602.

[0136] S903, determining the center line of the tubular object according to the initial center line.

[0137] Further, the computer device can determine the center line of the tubular object based on the initial center line. Optionally, the computer device can use the initial center line as the center line of the tubular object corresponding to the target connected domain, or can perform post-processing such as smoothing correction and outlier processing on the initial center line to determine the center line of the tubular object corresponding to the target connected domain, and this embodiment is not limited thereto.

[0138] This embodiment determines the first center line corresponding to the first projection plane based on the first projection plane, and determines the initial center line corresponding to the target connected domain based on the second projection plane and the first center line, and finally determines the center line of the tubular object based on the initial center line. Determining the attribute information of the tubular object on two orthogonal first projection planes and second projection planes greatly reduces the complexity of the fitting process, and at the same time, the extraction accuracy of the final target will not be greatly lost, and the fitting process can be completed quickly and efficiently. In addition, the center line of the tubular object is included in the attribute information, which is conducive to more accurate representation of the tubular object and facilitates subsequent three-dimensional reconstruction.

[0139] The target connected domain may have errors such as breakage and under-segmentation of the skull. Therefore, the computer device can search and optimize the initial center line to determine a more accurate center line of the tubular object. The following describes how to search and optimize the initial center line.

[0140] Fig.11 This is a schematic diagram of a process for determining a center line in an embodiment of the present application, referring to Fig.11 This embodiment relates to an optional implementation method of how to determine the center line of a tubular object. Based on the above embodiment, the above S903, determining the center line of the tubular object according to the initial center line, includes the following steps:

[0141] S1101, extending the first end and / or the second end of the initial center line in the region of interest to determine a first candidate center line.

[0142] In this embodiment, the target connected domain may be broken, the skull may be under-segmented, etc. Therefore, the computer device will extend at least one of the first end and the second end of the initial centerline within the scope of the region of interest to determine the first candidate centerline.

[0143] Optionally, the computer device may continue to calculate the initial centerline in the direction of the first end and / or the direction of the second end of the initial centerline using the RANSAC method within the scope of the region of interest to determine a first candidate centerline.

[0144] Fig.12 This is a schematic diagram of an extended process in an embodiment of the present application. Fig.12 As shown, Fig.12 The initial segmentation line 1201 corresponding to the target connected domain is included, and the first end of the initial segmentation line 1201 is extended in the region of interest to obtain a first candidate center line 1203.

[0145] S1102, determining a candidate search range according to the first candidate center line and the diameter of the tubular object.

[0146] Further, please continue to combine Fig.12 According to the first candidate centerline 1203 and the diameter of the tubular object, the candidate search range 1202 can be determined. It can be understood that according to the first candidate centerline 1203 and the radius of the tubular object, the candidate search range 1202 of the standard cylindrical shape can be determined; according to the first candidate centerline 1203 and the major axis length and minor axis length of the tubular object, the candidate search range 1202 of the elliptical cylinder shape can be determined.

[0147] S1103, determining a candidate connected domain according to pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold; the first preset threshold is determined according to each pixel value corresponding to the target connected domain.

[0148] In this embodiment, the computer device can determine the first preset threshold value according to the target connected domain. It can be understood that the target connected domain includes a plurality of pixel points in the region of interest of the medical image, and each pixel point corresponds to a pixel value. Taking the medical image as CT as an example, the pixel value of the pixel point is the CT value.

[0149] Optionally, the computer device may determine the first preset threshold value according to the pixel value of each pixel point in the target connected domain. For example, the computer device uses the 90% quantile of each pixel value as the first preset threshold value.

[0150] Furthermore, the computer device can use the first preset threshold to search for pixel points with pixel values ​​greater than the first preset threshold within the candidate search range. In other words, the computer device determines the candidate connected domain based on the pixel points with pixel values ​​greater than the first preset threshold in the candidate search range. Exemplarily, the computer device directly uses the pixel points with pixel values ​​greater than the first preset threshold in the candidate search range as the candidate connected domain.

[0151] Fig.13 is a schematic diagram of a candidate connected domain in an embodiment of the present application, such as Fig.13 As shown, the computer device can determine a candidate connected domain 1301, a candidate connected domain 1302, and a candidate connected domain 1303 according to the pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold.

[0152] S1104: Determine the center line of the tubular object according to the initial center line and the second candidate center line in the candidate connected domain.

[0153] based on Fig.13 The computer device can determine the center line of the tubular object corresponding to the target connected domain based on the initial center line 1201, the second candidate center line 1304 in the candidate connected domain 1301, the second candidate center line 1305 in the candidate connected domain 1302, and the second candidate center line 1306 in the candidate connected domain 1303.

[0154] For example, the computer device may connect the initial centerline 1201, the second candidate centerline 1304, the second candidate centerline 1305 and the second candidate centerline 1306 to obtain the centerline of the tubular object. It is understandable that the second candidate centerline belongs to the first candidate centerline.

[0155] In this embodiment, since the first end and / or the second end of the initial center line is extended in the region of interest to determine the first candidate center line, the missing of the initial center line can be compensated to a certain extent. Furthermore, since the first preset threshold is determined according to the pixel values ​​corresponding to the target connected domain, the candidate search range is determined according to the first candidate center line and the diameter of the tubular object, and the candidate search range is determined according to the first candidate center line and the diameter of the tubular object, and the candidate connected domain is determined according to the pixel points in the candidate search range whose pixel values ​​are greater than the first preset threshold, the tubular object that may be missed in the region of interest can be identified, thereby improving the recognition accuracy. Furthermore, according to the initial center line and the second candidate center line in the candidate connected domain, a more accurate center line of the tubular object can be determined.

[0156] Fig.14 This is another flow chart of determining the center line in the embodiment of the present application, referring to Fig.14 This embodiment relates to an optional implementation method of how to determine the center line of a tubular object. Based on the above embodiment, the above S1104, based on the initial center line and the second candidate center line in the candidate connected domain, determines the center line of the tubular object, including the following steps:

[0157] S1401, determining the distance between the initial center line and the adjacent center line; the adjacent center line is a second candidate center line adjacent to the initial center line.

[0158] In this embodiment, the computer device determines the initial center line and the distance between adjacent center lines. Fig.13 For example, the initial center line 1201 is adjacent to the second candidate center line 1304 , and the computer device determines the distance between the initial center line 1201 and the second candidate center line 1304 .

[0159] S1402, determining the center line of the tubular object according to the distance, the initial center line and the adjacent center lines.

[0160] In this embodiment, the computer device can determine the center line of the tubular object based on the distance, the initial center line and the adjacent center line. Optionally, when the distance is less than the preset distance, the computer device continues to calculate the distance between the adjacent center line and the next second candidate center line until the computer device calculates the last second candidate center line or calculates a distance not less than the preset distance. After that, the computer device can connect the second candidate center line with a distance less than the preset distance and the initial center line, and use the connected center line as the center line of the tubular object. The preset distance is a value set according to demand, which is a number greater than 0.

[0161] by Fig.13For example, if the distance between the initial center line 1201 and the second candidate center line 1304 is less than the preset distance, the computer device continues to calculate the distance between the second candidate center line 1304 and the adjacent second candidate center line 1305 .

[0162] If the distance between the second candidate centerline 1304 and the second candidate centerline 1305 is still less than the preset distance, the computer device continues to calculate the distance between the second candidate centerline 1305 and the adjacent second candidate centerline 1306. If the distance between the second candidate centerline 1305 and the second candidate centerline 1306 is not less than the preset distance, the computer device connects the initial centerline 1201, the second candidate centerline 1304 and the second candidate centerline 1305 to obtain the centerline of the tubular object. If the distance between the second candidate centerline 1305 and the second candidate centerline 1306 is less than the preset distance, the computer device connects the initial centerline 1201, the second candidate centerline 1304, the second candidate centerline 1305 and the second candidate centerline 1306 to obtain the centerline of the tubular object.

[0163] If the distance between the second candidate center line 1304 and the second candidate center line 1305 is not less than the preset distance, the computer device connects the initial center line 1201 and the second candidate center line 1304 to obtain the center line of the tubular object.

[0164] This embodiment determines the distance between the initial center line and the adjacent center line, and determines the center line of the tubular object based on the distance, the initial center line and the adjacent center line. Since the adjacent center line is a second candidate center line adjacent to the initial center line, the distance between the two adjacent center lines can be used to determine a relatively accurate center line of the tubular object.

[0165] Fig.15 This is another flow chart of determining the center line in the embodiment of the present application, referring to Fig.15 This embodiment relates to an optional implementation method of how to determine the center line of a tubular object. Based on the above embodiment, the above S1402, determining the center line of the tubular object according to the distance, the initial center line and the adjacent center line, includes the following steps:

[0166] S1501: If the distance is less than a preset distance, the initial center line and the adjacent center line are connected to update the initial center line.

[0167] In this embodiment, continue with Fig.13For example, during the first calculation, if the distance A between the initial center line 1201 and the second candidate center line 1304 is less than the preset distance, the computer device connects the initial center line 1201 and the second candidate center line 1304, and uses the result after connecting the initial center line 1201 and the second candidate center line 1304 as the new initial center line.

[0168] S1502, returning to the step of determining the distance between the initial center line and adjacent center lines until the distance is not less than a preset distance, and taking the initial center line when the distance is not less than the preset distance as the center line of the tubular object.

[0169] In this embodiment, continuing with the above example, during the second calculation, the initial centerline updates the result after the initial centerline 1201 and the second candidate centerline 1304 are connected, and the adjacent centerline is also updated to the second candidate centerline 1305. Therefore, the computer device calculates the result after the initial centerline 1201 and the second candidate centerline 1304 are connected, and the distance B between the second candidate centerline 1305.

[0170] Further, the computer device determines whether the distance B is less than a preset distance. If the distance B is not less than the preset distance, the computer device uses the result of connecting the initial center line 1201 and the second candidate center line 1304 as the center line of the tubular object.

[0171] If the distance B is less than the preset distance, the computer device will use the result after connecting the initial center line 1201, the second candidate center line 1304 and the second candidate center line 1305 as the new initial center line, and when entering the third calculation, continue to judge the result after connecting the initial center line 1201, the second candidate center line 1304 and the second candidate center line 1305, and the distance C between the second candidate center line 1306.

[0172] Afterwards, similarly, the computer device continues to determine the distance between the initial center line and adjacent center lines until the distance is not less than the preset distance, and uses the initial center line when the distance is not less than the preset distance as the center line of the tubular object. And so on, which will not be repeated here.

[0173] In this embodiment, if the distance is less than the preset distance, the initial center line and the adjacent center line are connected to update the initial center line, and then the step of determining the distance between the initial center line and the adjacent center line is returned to execute until the distance is not less than the preset distance, and the initial center line when the distance is not less than the preset distance is used as the center line of the tubular object. In this way, the accuracy of the center line of the tubular object is improved during continuous updating.

[0174] In one embodiment, optionally, if the distance between the initial center line and the adjacent center lines is less than a preset distance, the computer device may directly use the initial center line as the center line of the tubular object, or may obtain the center line of the tubular object after post-processing the initial center line.

[0175] In one embodiment, optionally, the above S202, determining the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain, can be implemented in the following manner:

[0176] An initial connected domain that meets preset conditions is used as a target connected domain; wherein the preset conditions include that a ratio between a first length corresponding to a first principal component and a second length corresponding to a second principal component is greater than a first multiple, and a ratio between the first length and a third length corresponding to a third principal component is greater than a second multiple, the first length is determined according to a coordinate range corresponding to the first principal component, the second length is determined according to a coordinate range corresponding to the second principal component, and the third length is determined according to a coordinate range corresponding to the third principal component.

[0177] In this embodiment, the length of the tubular object in the axial direction is much greater than its length in the radial direction. Therefore, according to this feature, the target connected domain can be determined from the initial connected domain using preset conditions.

[0178] Please continue to refer to Figure 5 According to the principal component analysis result of the initial connected domain, the computer device can determine the first length corresponding to the first principal component, the second length corresponding to the second principal component, and the third length corresponding to the third principal component. In other words, the first length is used to indicate the range size corresponding to each coordinate on the first principal component in the principal component analysis result, the second length is used to indicate the range size corresponding to each coordinate on the second principal component in the principal component analysis result, and the third length is used to indicate the range size corresponding to each coordinate on the third principal component in the principal component analysis result.

[0179] Furthermore, if the ratio between the first length and the second length is greater than the first multiple, and the ratio between the first length and the third length is greater than the second multiple, it means that the length of the initial connected domain in the axial direction is much greater than its length in the radial direction, which is consistent with the characteristics of a tubular object. Therefore, the initial connected domain can be used as a target connected domain.

[0180] The first multiple and the second multiple are numbers set according to demand, and can generally be set to 15 to 20. The first multiple and the second multiple can be the same or different.

[0181] Since the preset conditions include that the ratio of the first length corresponding to the first principal component to the second length corresponding to the second principal component is greater than the first multiple, and the ratio of the first length to the third length corresponding to the third principal component is greater than the second multiple, the first length is determined according to the coordinate range corresponding to the first principal component, the second length is determined according to the coordinate range corresponding to the second principal component, and the third length is determined according to the coordinate range corresponding to the third principal component. Therefore, the preset conditions can be used to screen out a target connected domain that better meets the characteristics of the tubular object from the initial connected domain, thereby improving the accuracy of the target connected domain.

[0182] Fig.16 This is a schematic diagram of a process for obtaining an initial connected domain in an embodiment of the present application, referring to Fig.16 , this embodiment relates to how to obtain an initial connected domain. An optional implementation method. Based on the above embodiment, the above medical image recognition method also includes the following steps:

[0183] S1601, obtaining medical images.

[0184] In this embodiment, the computer device can obtain medical images. The medical images may be sent by other devices or may be images stored in the computer device in advance, which is not limited in this embodiment.

[0185] In some embodiments, after acquiring the initial medical image, the computer device may pre-process the initial medical image to obtain a medical image for subsequent use. The pre-processing may include but is not limited to correction processing, cropping processing, and the like.

[0186] S1602, segment the medical image to obtain a region of interest.

[0187] Optionally, the computer device may perform threshold segmentation on the medical image to obtain the region of interest, or may input the medical image into a trained segmentation model to obtain the region of interest in the medical image. The segmentation model may include but is not limited to convolutional neural networks (CNN), recurrent neural networks (RNN), or other deep learning networks, machine learning networks, etc.

[0188] Exemplarily, the computer device performs threshold segmentation on the CT image of the brain, removes the skull in the medical image, and retains only the CT brain tissue as the region of interest.

[0189] S1603: Perform threshold segmentation on the region of interest according to a second preset threshold to obtain an initial connected region.

[0190] Furthermore, in medical images, tissues and tubular objects have a greater degree of distinction in pixel values, so the second preset threshold can be used to perform threshold segmentation on the region of interest to obtain an initial connected domain. Exemplarily, the computer device performs a connected domain analysis on each pixel point in the region of interest whose pixel value is greater than the second preset threshold to obtain an initial connected domain.

[0191] This embodiment acquires a medical image, segments the medical image to obtain a region of interest, and then performs threshold segmentation on the region of interest according to a second preset threshold to obtain an initial connected domain, thereby improving the efficiency and accuracy of obtaining the initial connected domain.

[0192] Fig.17 This is a schematic diagram of a process for displaying reconstruction results in an embodiment of the present application, referring to Fig.17 This embodiment relates to an optional implementation method of how to reconstruct the result. Based on the above embodiment, the above medical image recognition method also includes the following steps:

[0193] S1701, modeling the tubular object according to the attribute information to obtain a reconstruction result of the tubular object.

[0194] S1702, displaying the reconstruction result of the tubular object in the medical image.

[0195] In this embodiment, after acquiring the attribute information, the computer device can model the tubular object according to the attribute information to obtain a reconstruction result of the tubular object. Fig.18 is a schematic diagram of a reconstruction result in an embodiment of the present application, such as Fig.18 As shown, Fig.18 A case where the tubular object is a standard cylinder is shown, and r represents the radius of the standard cylinder, which is determined according to the diameter of the standard cylinder.

[0196] Furthermore, the computer device can display the reconstruction result of the tubular object in the medical image. For example, the computer device displays the reconstruction result of the tubular object at the corresponding position of the tubular object in the medical image to facilitate the doctor's identification.

[0197] This embodiment models the tubular object according to the attribute information, obtains the reconstruction result of the tubular object, and displays the reconstruction result of the tubular object in the medical image, so that the doctor can quickly and conveniently identify the tubular object in the medical image.

[0198] In order to more clearly introduce the medical image recognition method in this application, Fig.19 and Fig. 20 Provide explanation. Fig.19 is a diagram illustrating the principle of a medical image recognition method in an embodiment of the present application, Fig. 20 This is a process diagram of a medical image recognition method in an embodiment of the present application. Fig.19 and Fig. 20 , the computer device can execute the above-mentioned medical image recognition method according to the following process.

[0199] S2001, obtaining medical images.

[0200] S2002, segment the medical image to obtain the region of interest. Fig.19 The steps of extracting the region of interest in .

[0201] S2003, performing threshold segmentation on the region of interest according to a second preset threshold to obtain an initial connected domain. Fig.19 The step of extracting the initial connected domain.

[0202] S2004, according to the principal component analysis results corresponding to the initial connected domain, the initial connected domain that meets the preset conditions is used as the target connected domain. The preset conditions include that the ratio between the first length corresponding to the first principal component and the second length corresponding to the second principal component is greater than the first multiple, and the ratio between the first length and the third length corresponding to the third principal component is greater than the second multiple, the first length is determined according to the coordinate range corresponding to the first principal component, the second length is determined according to the coordinate range corresponding to the second principal component, and the third length is determined according to the coordinate range corresponding to the third principal component. S2001 to S2004 are Fig.19 Step 2004: determining the target connected domain. Fig.19 That is to say, after principal component analysis, the target connected domain can be determined.

[0203] S2005, determining a first projection surface and a second projection surface according to the principal component analysis result corresponding to the target connected domain. The first projection surface is determined according to the first principal component and the second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to the first principal component and the third principal component in the principal component analysis result corresponding to the target connected domain;

[0204] S2006, for each first candidate coordinate value, determine a first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface, and determine the first tube diameter according to the first difference corresponding to each first candidate coordinate value.

[0205] S2007, for each second candidate coordinate value, determine a second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface, and determine the second tube diameter according to the second difference corresponding to each second candidate coordinate value.

[0206] S2008, if the difference between the first tube diameter and the second tube diameter is greater than a preset difference, determining the major axis length of the tubular object according to the maximum value of the first tube diameter and the second tube diameter, and determining the minor axis length of the tubular object according to the minimum value of the first tube diameter and the second tube diameter.

[0207] S2009, if the difference between the first tube diameter and the second tube diameter is not greater than the preset difference, the diameter of the tubular object is determined according to the average value of the first tube diameter and the second tube diameter. Fig.19 Steps for confirming the pipe diameter.

[0208] S2010: Based on the first projection surface, determine a first center line corresponding to the first projection surface.

[0209] S2011, based on the second projection surface and the first center line, determine the initial center line corresponding to the target connected domain. Fig.19 The steps to confirm the initial center line. Fig.19 Roughly extracted steps.

[0210] S2012: Extend the first end and / or the second end of the initial centerline in the region of interest to determine a first candidate centerline.

[0211] S2013, determining a candidate search range according to the first candidate center line and the diameter of the tubular object. Fig.19 Steps for confirming the candidate search range in .

[0212] S2014, determining a candidate connected domain according to the pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold. The first preset threshold is determined according to the pixel values ​​corresponding to the target connected domain. Fig.19 The step of extracting candidate connected components.

[0213] S2015: Determine the distance between the initial center line and the adjacent center line, wherein the adjacent center line is a second candidate center line adjacent to the initial center line.

[0214] S2017: If the distance is not less than the preset distance, the initial center line is used as the center line of the tubular object.

[0215] S2016: If the distance is less than a preset distance, the initial center line and the adjacent center line are connected to update the initial center line.

[0216] S2018, returning to the step of determining the distance between the initial center line and adjacent center lines until the distance is not less than the preset distance, and taking the initial center line when the distance is not less than the preset distance as the center line of the tubular object. Fig.19 The steps for determining the center line in S2012 to S2018 are Fig.19 Steps in search optimization.

[0217] S2019: Model the tubular object according to the attribute information to obtain a reconstruction result of the tubular object.

[0218] S2020, the reconstruction results of tubular objects are displayed in medical images. Fig.19 Steps for identifying tubular objects.

[0219] S2001 to S2020 may refer to the above embodiment and will not be described in detail here.

[0220] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.

[0221] Based on the same inventive concept, the embodiment of the present application also provides a medical image recognition device for implementing the medical image recognition method involved above. The implementation solution provided by the device to solve the problem is similar to the implementation solution recorded in the above method, so the specific limitations in one or more medical image recognition device embodiments provided below can refer to the limitations of the medical image recognition method above, and will not be repeated here.

[0222] Fig.21 is a structural block diagram of a medical image recognition device in an embodiment of the present application, such as Fig.21 As shown, in an embodiment of the present application, a medical image recognition device 2100 is provided, including: a first determination module 2101, a second determination module 2102 and a third determination module 2103, wherein:

[0223] The first determination module 2101 is used to perform principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain a principal component analysis result corresponding to the initial connected domain;

[0224] A second determination module 2102 is used to determine a target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain;

[0225] The third determination module 2103 is used to determine the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain.

[0226] The medical image recognition device provided in this embodiment performs principal component analysis on the initial connected domain of the tubular object in the region of interest of the medical image to obtain the principal component analysis result corresponding to the initial connected domain, and determines the target connected domain from the initial connected domain according to the principal component analysis result corresponding to the initial connected domain, and then determines the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain. Since the initial connected domain can be screened according to the principal component analysis result corresponding to the initial connected domain, a more accurate target connected domain of the tubular object can be determined from the initial connected domain. Furthermore, after the attribute information of the tubular object is determined according to the principal component analysis result of the target connected domain, the shape, position and other information of the tubular object can be determined, thereby realizing efficient and accurate recognition of the tubular object from the medical image.

[0227] Optionally, the attribute information includes the diameter of the tubular object; the third determination module 2103 includes:

[0228] A first determining unit is used to determine a first projection surface and a second projection surface according to a principal component analysis result corresponding to the target connected domain; the first projection surface is determined according to a first principal component and a second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to a first principal component and a third principal component in the principal component analysis result corresponding to the target connected domain;

[0229] A second determination unit is used to determine a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of the first principal component in the first projection surface, and to determine a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface;

[0230] The third determining unit is used to determine the diameter of the tubular object according to the first diameter and the second diameter.

[0231] Optionally, the second determining unit includes:

[0232] The first determination subunit is used to determine the first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface for each first candidate coordinate value, and determine the first tube diameter according to the first difference corresponding to each first candidate coordinate value.

[0233] The second determination subunit is used to determine the second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface for each second candidate coordinate value, and determine the second pipe diameter according to the second difference corresponding to each second candidate coordinate value.

[0234] Optionally, the third determining unit includes:

[0235] The third determination subunit is used to determine the major axis length of the tubular object according to the maximum value of the first tube diameter and the second tube diameter if the difference between the first tube diameter and the second tube diameter is greater than a preset difference, and to determine the minor axis length of the tubular object according to the minimum value of the first tube diameter and the second tube diameter.

[0236] The fourth determining subunit is used to determine the diameter of the tubular object according to an average value of the first tube diameter and the second tube diameter if the difference between the first tube diameter and the second tube diameter is not greater than a preset difference.

[0237] Optionally, the attribute information includes a center line of the tubular object; the third determination module 2103 includes:

[0238] The fourth determining unit is used to determine a first center line corresponding to the first projection surface based on the first projection surface.

[0239] The fifth determining unit is used to determine an initial center line corresponding to the target connected domain based on the second projection surface and the first center line.

[0240] The sixth determining unit is used to determine the center line of the tubular object according to the initial center line.

[0241] Optionally, the sixth determining unit includes:

[0242] The fifth determination subunit is used to extend the first end and / or the second end of the initial center line in the region of interest to determine a first candidate center line.

[0243] The sixth determination subunit is used to determine a candidate search range according to the first candidate center line and the diameter of the tubular object.

[0244] A seventh determination subunit is used to determine a candidate connected domain according to pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold; the first preset threshold is determined according to each pixel value corresponding to the target connected domain;

[0245] The eighth determination subunit is used to determine the center line of the tubular object according to the initial center line and the second candidate center line in the candidate connected domain.

[0246] Optionally, the eighth determination subunit is also used to determine the distance between the initial center line and the adjacent center line; the adjacent center line is a second candidate center line adjacent to the initial center line; and the center line of the tubular object is determined based on the distance, the initial center line and the adjacent center line.

[0247] Optionally, the eighth determination subunit is also used to connect the initial center line and the adjacent center line to update the initial center line if the distance is less than a preset distance; return to execute the step of determining the distance between the initial center line and the adjacent center line until the distance is not less than a preset distance, and use the initial center line when the distance is not less than the preset distance as the center line of the tubular object.

[0248] Optionally, the second determining module 2102 is further configured to use an initial connected domain that meets a preset condition as a target connected domain;

[0249] Among them, the preset conditions include that the ratio of a first length corresponding to the first principal component to the second length corresponding to the second principal component is greater than a first multiple, and the ratio of the first length to the third length corresponding to the third principal component is greater than a second multiple, the first length is determined according to the coordinate range corresponding to the first principal component, the second length is determined according to the coordinate range corresponding to the second principal component, and the third length is determined according to the coordinate range corresponding to the third principal component.

[0250] Optionally, the medical image recognition device 2100 includes:

[0251] An acquisition module, used for acquiring medical images;

[0252] A first segmentation module is used to segment the medical image to obtain a region of interest;

[0253] The second segmentation module performs threshold segmentation on the region of interest according to a second preset threshold to obtain an initial connected domain.

[0254] Optionally, the medical image recognition device 2100 includes:

[0255] A modeling module, used for modeling the tubular object according to the attribute information to obtain a reconstruction result of the tubular object;

[0256] The display module is used to display the reconstruction result of the tubular object in the medical image.

[0257] Each module in the above-mentioned medical image recognition device can be implemented in whole or in part by software, hardware or a combination thereof. Each of the above-mentioned modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute the operations corresponding to each of the above modules.

[0258] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above-mentioned method embodiments when executing the computer program.

[0259] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps in the above-mentioned method embodiments are implemented.

[0260] In one embodiment, a computer program product is provided, including a computer program, which implements the steps in the above method embodiments when executed by a processor.

[0261] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in the present application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.

[0262] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0263] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A medical image recognition method, characterized in that: The method comprises: Performing principal component analysis on an initial connected domain of a tubular object in a region of interest of a medical image to obtain a principal component analysis result corresponding to the initial connected domain; Determining a target connected domain from the initial connected domain according to a principal component analysis result corresponding to the initial connected domain; The attribute information of the tubular object is determined according to the principal component analysis result corresponding to the target connected domain.

2. The method according to claim 1, characterized in that The attribute information includes the diameter of the tubular object; The determining of attribute information according to the principal component analysis result corresponding to the target connected domain includes: Determine a first projection surface and a second projection surface according to a principal component analysis result corresponding to the target connected domain; the first projection surface is determined according to a first principal component and a second principal component in the principal component analysis result corresponding to the target connected domain, and the second projection surface is determined according to a first principal component and a third principal component in the principal component analysis result corresponding to the target connected domain; Determine a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of the first principal component in the first projection surface, and determine a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface; The diameter of the tubular object is determined according to the first diameter and the second diameter.

3. The method according to claim 2, characterized in that The method of determining a first tube diameter corresponding to the first projection surface according to a plurality of first candidate coordinate values ​​of the first principal component in the first projection surface, and determining a second tube diameter corresponding to the second projection surface according to a plurality of second candidate coordinate values ​​of the first principal component in the second projection surface, comprises: For each of the first candidate coordinate values, determine a first difference between the minimum coordinate value and the maximum coordinate value of the second principal component in the first projection surface, and determine the first tube diameter according to the first difference corresponding to each of the first candidate coordinate values; and For each of the second candidate coordinate values, determine the second difference between the minimum coordinate value and the maximum coordinate value of the third principal component in the second projection surface, and determine the second tube diameter according to the second difference corresponding to each of the second candidate coordinate values.

4. The method according to claim 3, characterized in that The step of determining the diameter of the tubular object according to the first diameter and the second diameter includes: If the difference between the first tube diameter and the second tube diameter is greater than a preset difference, determining the major axis length of the tubular object according to the maximum value of the first tube diameter and the second tube diameter, and determining the minor axis length of the tubular object according to the minimum value of the first tube diameter and the second tube diameter; If the difference between the first tube diameter and the second tube diameter is not greater than the preset difference, the diameter of the tubular object is determined according to an average value of the first tube diameter and the second tube diameter.

5. The method according to any one of claims 2 to 4, characterized in that: The attribute information includes the center line of the tubular object; and determining the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain includes: Based on the first projection surface, determining a first center line corresponding to the first projection surface; Determining an initial center line corresponding to the target connected domain based on the second projection surface and the first center line; The centerline of the tubular object is determined based on the initial centerline.

6. The method according to claim 5, characterized in that Determining the center line of the tubular object according to the initial center line includes: Extending the first end and / or the second end of the initial centerline in the region of interest to determine a first candidate centerline; Determining a candidate search range according to the first candidate center line and the diameter of the tubular object; Determine a candidate connected domain according to pixel points in the candidate search range whose pixel values ​​are greater than a first preset threshold; the first preset threshold is determined according to each pixel value corresponding to the target connected domain; The center line of the tubular object is determined according to the initial center line and the second candidate center line in the candidate connected domain.

7. The method according to claim 6, characterized in that The step of determining the center line of the tubular object according to the initial center line and the second candidate center line in the candidate connected domain comprises: Determine the distance between the initial center line and an adjacent center line; the adjacent center line is a second candidate center line adjacent to the initial center line; The center line of the tubular object is determined according to the distance, the initial center line and the adjacent center lines.

8. The method according to claim 7, characterized in that Determining the center line of the tubular object according to the distance, the initial center line and the adjacent center line comprises: If the distance is less than a preset distance, connecting the initial center line and the adjacent center line to update the initial center line; Return to the step of determining the initial center line and the distance between adjacent center lines until the distance is not less than the preset distance, and use the initial center line when the distance is not less than the preset distance as the center line of the tubular object.

9. The method according to any one of claims 2 to 4, characterized in that: The step of determining a target connected domain from the initial connected domain according to a principal component analysis result corresponding to the initial connected domain comprises: Taking the initial connected domain that meets the preset conditions as the target connected domain; Among them, the preset conditions include that the ratio of a first length corresponding to the first principal component to a second length corresponding to the second principal component is greater than a first multiple, and the ratio of the first length to a third length corresponding to the third principal component is greater than a second multiple, the first length is determined according to the coordinate range corresponding to the first principal component, the second length is determined according to the coordinate range corresponding to the second principal component, and the third length is determined according to the coordinate range corresponding to the third principal component.

10. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: acquiring the medical image; Segmenting the medical image to obtain the region of interest; The region of interest is segmented by a threshold value according to a second preset threshold value to obtain the initial connected domain.

11. The method according to any one of claims 1 to 4, characterized in that: The method further comprises: Modeling the tubular object according to the attribute information to obtain a reconstruction result of the tubular object; The reconstruction result of the tubular object is displayed in the medical image.

12. A medical image recognition device, characterized in that: The device comprises: A first determination module is used to perform principal component analysis on an initial connected domain of tubular objects in a region of interest of a medical image to obtain a principal component analysis result corresponding to the initial connected domain; A second determination module is used to determine a target connected domain from the initial connected domain according to a principal component analysis result corresponding to the initial connected domain; The third determination module is used to determine the attribute information of the tubular object according to the principal component analysis result corresponding to the target connected domain.

13. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 11 are implemented.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 11 are implemented.

Images