Method and device for optimizing three-dimensional large intestine segmentation and storage medium

Abstract

The invention discloses a method and equipment for optimizing three-dimensional large intestine segmentation and a storage medium. The method comprises the following steps: obtaining a three-dimensional large intestine segmentation result, wherein the three-dimensional large intestine segmentation result comprises a plurality of connected blocks; extracting an end point of the center line of each connected block, and constructing a target distance function based on the end point of the center line of each connected block; and minimizing the target distance function under a target constraint condition to enable the plurality of connected blocks to form a single connected block so as to optimize the three-dimensional large intestine segmentation result. By means of the scheme, the large intestine segmentation results can be communicated, it is ensured that the large intestine segmentation results pass through all points, the length is minimum, and convenience is provided for follow-up analysis.

Description

Technical Field

[0001] This application generally relates to the field of image processing technology. More specifically, this application relates to a method, device, and computer-readable storage medium for optimizing three-dimensional large intestine segmentation. Background Art

[0002] With the development of medical image processing and three-dimensional visualization technology, virtual endoscopy has been widely used in many clinical experiments and various medical diagnoses, such as intestinal examinations, due to its obvious advantages such as non-invasiveness and repeatability. Virtual endoscopy first obtains tomographic scan data of the human body through medical scanning devices such as computed tomography ("CT") and magnetic resonance imaging ("MRI"), and then uses image processing technology to reconstruct three-dimensional images to form virtual human tissues.

[0003] In intestinal examinations, in addition to the large intestine part, other tissue organs may also exist in the reconstructed three-dimensional intestinal image, which will interfere with subsequent research. Therefore, it is necessary to segment the large intestine from the three-dimensional image. However, due to incomplete inflation of the large intestine or excessive fluid accumulation inside the large intestine when taking the three-dimensional image, the final large intestine segmentation result is not simply connected, affecting subsequent analysis (such as conformal unfolding or three-dimensional navigation, etc.). Therefore, how to optimize the large intestine segmentation result to make it simply connected has become a technical problem that needs to be solved urgently. Currently, the persistent homology method or the shortest path method has been used to achieve simple connectivity. However, the execution method of persistent homology is essentially a greedy strategy, which only considers the nearest point of each point separately, without considering whether the finally formed loop is the overall shortest. Although the shortest path algorithm is a global shortest path, it cannot guarantee that the shortest path contains every point. According to anatomical knowledge, a complete large intestine must obviously pass through all points and have the minimum length.

[0004] In view of this, there is an urgent need to provide a solution for optimizing three-dimensional large intestine segmentation. By constructing an objective distance function based on the endpoints of the lines in each connected component of the three-dimensional large intestine segmentation result, and minimizing the objective distance function through the objective constraint conditions set based on 0-1 integer programming, multiple connected components in the three-dimensional large intestine segmentation result are made into a simply connected component, ensuring that all points are passed through and the length is minimized, providing convenience for subsequent analysis. Summary of the Invention

[0005] In order to solve at least one or more of the above-mentioned technical problems, this application proposes a solution for optimizing three-dimensional large intestine segmentation in multiple aspects.

[0006] In a first aspect, the present application provides a method for optimizing three-dimensional large intestine segmentation, including: obtaining a three-dimensional large intestine segmentation result, where the three-dimensional large intestine segmentation result includes a plurality of connected components; extracting the endpoints of the midlines of each of the connected components, and constructing an objective distance function based on the endpoints of the midlines of each of the connected components; and minimizing the objective distance function under objective constraint conditions set based on 0-1 integer programming, such that the plurality of connected components form a single connected component to optimize the three-dimensional large intestine segmentation result.

[0007] In one embodiment, constructing an objective distance function based on the endpoints of the midlines of each of the connected components includes: setting objective variables based on the endpoints of the midlines of each of the connected components; and constructing the objective distance function based on the product of the objective variables and the distances between the endpoints of the midlines of each of the connected components.

[0008] In another embodiment, the objective distance function is constructed by the following formula: , where represents the distance between the endpoints of the midlines of each of the connected components, represents the objective variable, and i, j represent the endpoints of the midlines of the connected components.

[0009] In yet another embodiment, minimizing the objective distance function under objective constraint conditions set based on 0-1 integer programming, such that the plurality of connected components form a single connected component to optimize the three-dimensional large intestine segmentation result includes: determining target endpoints by minimizing the objective distance function under the objective constraint conditions set based on 0-1 integer programming; and connecting all the target endpoints such that the plurality of connected components form a single connected component to optimize the three-dimensional large intestine segmentation result.

[0010] In yet another embodiment, the objective constraint conditions set based on 0-1 integer programming include: setting the objective variables corresponding to the objective distances to be summed in the objective distance function to a first objective value, and setting the objective variables corresponding to the remaining objective distances to a second objective value; restricting the values of the objective variables such that each of the connected components is connected to only one straight line segment; and restricting the values of the target endpoints to avoid adding redundant straight line segments.

[0011] In yet another embodiment, the first objective value is 1 and the second objective value is 0.

[0012] In yet another embodiment, the values of the objective variables are restricted by the following formula such that each of the connected components is connected to only one straight line segment: and , where represents that only one straight line is connected to endpoint j, It means that only one straight line is connected to the endpoint i.

[0013] In yet another embodiment, the value of the target endpoint is restricted by the following formula to avoid adding unnecessary line segments: , where represents the line segment connecting the starting point and the ending point, represents the existing median line.

[0014] In a second aspect, the present application provides an apparatus for optimizing three-dimensional large intestine segmentation, including: a processor; and a memory storing computer instructions for optimizing three-dimensional large intestine segmentation, which, when executed by the processor, implement one or more embodiments in the foregoing first aspect.

[0015] In a third aspect, the present application provides a computer-readable storage medium storing computer program instructions for optimizing three-dimensional large intestine segmentation, which, when executed by one or more processors, implement one or more embodiments in the foregoing first aspect.

[0016] Through the solution for optimizing three-dimensional large intestine segmentation provided above, embodiments of the present application construct a target distance function based on the endpoints of the median lines of each connected component in the three-dimensional large intestine segmentation result, and minimize the target distance function through the target constraint conditions set based on 0-1 integer programming, so that multiple connected components in the three-dimensional large intestine segmentation result form a single connected component, ensuring that all points are passed through and the length is minimized to complete the optimization of the three-dimensional large intestine segmentation result for subsequent analysis. In some embodiments, the foregoing target constraint conditions include setting corresponding target values for the target variables corresponding to the target distances required to be summed and the target variables not to be summed in the target distance function, restricting the values of the target variables and the values of the target endpoints. This makes the straight-line distance between the target endpoints the shortest, the total straight-line length the shortest, and each connected region is only connected to one straight-line segment and does not connect to the first and last two endpoints (i.e., the cecal end and the rectal end), so as to connect multiple connected components in the three-dimensional large intestine segmentation result into a single connected component. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features, and advantages of the exemplary embodiments of the present application will become readily understood. In the drawings, several embodiments of the present application are shown in an exemplary rather than restrictive manner, and the same or corresponding reference numerals represent the same or corresponding parts, where: Figure 1 is an exemplary flowchart showing a method for optimizing three-dimensional large intestine segmentation according to an embodiment of the present application; Figure 2 is an exemplary schematic diagram showing a three-dimensional large intestine segmentation result according to an embodiment of the present application; Figure 3 It is an exemplary schematic diagram showing the midlines and endpoints of each connected component in the three-dimensional large intestine segmentation result according to an embodiment of the present application; Figure 4 It is an exemplary schematic diagram showing connecting multiple connected components into a single connected component according to an embodiment of the present application; Figure 5 It is an exemplary structural block diagram of a device for optimizing three-dimensional large intestine segmentation according to an embodiment of the present application. Detailed implementation manners

[0018] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts shall fall within the protection scope of the present application.

[0019] It should be understood that the terms "include" and "comprise" used in the specification and claims of the present application indicate the presence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0020] It should also be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and claims of the present application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in the specification and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0021] As used in the present specification and claims, the term "if" can be interpreted as "when...", "once", "in response to determining", or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if detecting [the described condition or event]" can be interpreted as meaning "once determined", "in response to determining", "once detecting [the described condition or event]", or "in response to detecting [the described condition or event]" according to the context.

[0022] Next, the detailed implementation manners of the present application will be described in detail in conjunction with the accompanying drawings.

[0023] Figure 1is an exemplary flowchart showing a method 100 for optimizing three-dimensional large intestine segmentation according to an embodiment of the present application. As Figure 1 shown, at step S101, a three-dimensional large intestine segmentation result is obtained. In one implementation scenario, the three-dimensional large intestine segmentation result can be obtained by performing large intestine segmentation on a three-dimensional intestinal image. In some embodiments, the large intestine can be segmented from the three-dimensional intestinal image by methods such as threshold segmentation, edge segmentation, or deep learning to obtain the three-dimensional large intestine segmentation result. As mentioned above, due to incomplete inflation of the large intestine or excessive internal fluid accumulation in the large intestine when taking the three-dimensional image, the final large intestine segmentation result is not simply connected, resulting in multiple connected components (e.g., Figure 2 shown with four connected components), causing discontinuous jumps when randomly walking inside the large intestine to simulate virtual colonoscopy. To connect the large intestine, an effective method is to add some straight line segments, and the added straight line segments need to have the following properties: the added straight line segments should be the shortest paths between two connected components; the total length of the added straight line segments is the shortest; each connected region can and can only be connected to one straight line segment; and the cecum end (the connection with the small intestine) and the rectum end (the anus) cannot be connected.

[0024] Specifically, based on the three-dimensional large intestine segmentation result including multiple connected components, at step S102, the endpoints of the midlines of each connected component are extracted, and a target distance function is constructed based on the endpoints of the midlines of each connected component. It can be understood that the endpoints of the midlines of each connected component mentioned above include the starting point and the ending point of the midline. In one implementation scenario, first, each connected component can be parameterized into a two-dimensional rectangular domain, and then multiple groups of target point sets are collected along the longitudinal direction of each rectangular domain, and each group of target point sets contains multiple target points. Calculate the coordinate mean of each group of multiple target points, and the point corresponding to the coordinate mean is the mid-axis point of each connected component. It can be understood that the multiple groups of target point sets correspond to multiple mid-axis points. By mapping the multiple mid-axis points back to each connected component and connecting them, the midlines of each connected component can be extracted (e.g., Figure 3 shown), and then the endpoints of the midlines of each connected component can be obtained. By selecting the two endpoints of the midlines of each connected component, the path of the straight line segments added between each connected component can be made the shortest in the subsequent process, and by selecting the two endpoints of the midline, it can also be avoided that forks and loops appear in the subsequent straight line addition. In some embodiments, the aforementioned parameterization can be implemented by methods such as conformal mapping or area-preserving mapping.

[0025] Further, a target distance function is constructed based on the endpoints of the midlines of each connected component. In one embodiment, target variables can be set based on the endpoints of the midlines of each connected component, and the target distance function can be constructed based on the product of the target variables and the distances between the endpoints of the midlines of each connected component. It can be understood that the aforementioned target distance function is the sum of the Euclidean distances of the newly added line segments. Specifically, the target distance function can be constructed by the following formula: , where represents the distance between the endpoints of the midlines of each connected component, represents the target variable, and i, j represent the endpoints of the midlines of the connected components. It should be understood that for connected components, there are endpoints, and there are line segments. Therefore, target variables can be set, and . As an example, assume that there are two endpoints i and j, and the possible line segments they may have include i-i, i-j, j-i, j-j. The aforementioned target variable represents whether there is a line segment between endpoint i and endpoint j, and its value is 0 or 1. 1 means there is a line segment between endpoint i and endpoint j, and 0 means there is no line segment between endpoint i and endpoint j.

[0026] According to the target distance function constructed above, at step S103, the target distance function is minimized under the target constraint conditions set based on 0-1 integer programming, so that multiple connected components form a single connected component to optimize the three-dimensional large intestine segmentation result. In one embodiment, the target endpoints are determined by minimizing the target distance function under the target constraint conditions set based on 0-1 integer programming, and multiple connected components form a single connected component by connecting all the target endpoints to optimize the three-dimensional large intestine segmentation result. That is, the target endpoints connecting the connected domains are determined by minimizing the target distance function, and line segments are added between the target endpoints to connect the multiple connected components, so that the multiple connected components in the three-dimensional large intestine segmentation result are connected into a single connected component.

[0027] In one implementation scenario, the above target constraint conditions may include setting the target variables corresponding to the target distances required to be summed in the target distance function to a first target value, and setting the target variables corresponding to the remaining target distances to a second target value, restricting the values of the target variables so that each connected component is only connected to one line segment, and restricting the values of the target endpoints to avoid adding unnecessary line segments. Among them, the first target value is 1, and the second target value is 0. That is, the corresponding to in the summation is set to , and the corresponding to not in the summation is set to , so that the total length of the added straight line segments is the shortest.

[0028] In one implementation scenario, the value of the target variable can be restricted by the following formula so that each connected component is only connected to one straight line segment: and , where means that only one straight line is connected to endpoint j, means that only one straight line is connected to endpoint i. This enables the construction of a path without loops and bifurcations. According to anatomical knowledge, the complete large intestine clearly meets this requirement. In another implementation scenario, the value of the target endpoint can be restricted by the following formula to avoid adding unnecessary straight line segments: , where represents the straight line segment connecting the starting point and the ending point, represents the existing median line. This is because the end of the cecum (where it connects to the small intestine) and the end of the rectum (at the anus) cannot be connected, and setting to 0 can meet the requirement. In some embodiments, there may be solutions, that is, a connection is made between endpoint and endpoint , while there is already a median line between endpoint and endpoint . To avoid adding unnecessary straight line segments, set .

[0029] As an example, minimizing the target distance function under the above target constraint conditions can be expressed by the following formula: , where , , . Additionally, the corresponding to the sum is set to for , and the corresponding to those not in the sum is set to for . Under the above multiple constraint conditions, by solving, for example, using the branch and bound method , all target endpoints can be obtained. Further, by connecting all the target endpoints, a complete median line can be obtained to achieve the simple connectivity of the large intestine.

[0030] As described above, in the embodiment of the present application, a target distance function is constructed based on the end points of the lines in each connected component of the three-dimensional large intestine segmentation result, and the target distance function is minimized through the target constraint conditions set based on 0-1 integer programming, so that multiple connected components in the three-dimensional large intestine segmentation result form a single connected component, completing the optimization of the three-dimensional large intestine segmentation result and facilitating subsequent analysis. For example, virtual colonoscopy can be simulated within the single-connected large intestine for random walking and polyp observation. Further, through multiple target constraint conditions, the straight-line distance between the target end points is the shortest, the total direct length is the shortest, and each connected region is connected to only one straight-line segment and does not connect the first and last two end points, so as to connect multiple connected components in the three-dimensional large intestine segmentation result into a single connected component.

[0031] Figure 2 is an exemplary schematic diagram showing the three-dimensional large intestine segmentation result according to the embodiment of the present application. As Figure 2 shown in the figure, the three-dimensional large intestine segmentation result includes four connected components, namely connected component A, connected component B, connected component C, and connected component D. According to the foregoing, the large intestine can be segmented from the three-dimensional intestinal image by methods such as threshold segmentation, edge segmentation, or deep learning to obtain the three-dimensional large intestine segmentation result. Among them, threshold segmentation refers to classifying the gray histogram of the image by setting a threshold, and allocating pixels to different categories according to the comparison result of their gray values with the threshold to extract the large intestine segmentation result. The foregoing threshold segmentation methods include, for example, the histogram bimodal method, the adaptive threshold method, etc. Edge segmentation is to segment by detecting the edges in the image to obtain the large intestine segmentation result. The foregoing edge segmentation algorithms include, for example, the Sobel algorithm, the Laplacian algorithm, etc. For the deep learning method, a trained segmentation model can be used to segment the large intestine to obtain the large intestine segmentation result. In addition, the three-dimensional large intestine segmentation result can also be obtained by methods such as graph segmentation, clustering segmentation, or region growing, and the present application does not make any restrictions in this regard.

[0032] Figure 3 is an exemplary schematic diagram showing the midlines and end points of each connected component in the three-dimensional large intestine segmentation result according to the embodiment of the present application. As Figure 3 shown in the figure, the midlines corresponding to connected component A, connected component B, connected component C, and connected component D are l1, l2, l3, and l4 respectively. As an example, the two end points of the midline l1 are denoted as and , the two end points of the midline l2 are denoted as and , the two end points of the midline l3 are denoted as and and the two end points of the midline l4 are denoted as and . For n connected components, the two end points of the i-th midline can be denoted as and , wherein is the end of the rectum, and is at the end of the cecum.

[0033] In one embodiment, first, the target variables can be set according to the endpoints of the midlines of each connected component , and based on the target variables and the distances between the endpoints of the midlines of each connected component a target distance function is constructed by taking the product . Then, by minimizing the target distance function under the target constraints. Specifically, the foregoing target constraints include setting the corresponding to the sum to be , and the corresponding to those not in the sum is set to . Thus, the total length of the added straight line segments can be made the shortest. Further, the target constraints further include setting , , , so that each connected region is only connected to one straight line segment and the two endpoints of the cecum end and the rectum end are not connected, so as to connect the multiple connected components in the three-dimensional large intestine segmentation result into a single connected component.

[0034] Figure 4 is an exemplary schematic diagram showing connecting multiple connected components into a single connected one according to an embodiment of the present application. As Figure 4 shown in the figure is a schematic diagram of connecting multiple connected components into a single connected one. It can be seen from the figure that the midlines of the multiple connected components are connected into one midline, completing the connection of the large intestine in the large intestine segmentation result, thus facilitating subsequent analysis.

[0035] Figure 5 is an exemplary structural block diagram of a device 500 for optimizing three-dimensional large intestine segmentation according to an embodiment of the present application. It can be understood that the device 500 may include the device of the embodiment of the present application, and the device implementing the solution of the present application may be a single device (such as a computing device) or a multifunctional device including various peripheral devices.

[0036] As Figure 5As shown, the device of the present application may further include a central processing unit or central processing unit ("CPU") 511, which may be a general-purpose CPU, a dedicated CPU, or other information processing and program execution units. Further, the device 500 may also include a large-capacity memory 512 and a read-only memory ("ROM") 513, where the large-capacity memory 512 may be configured to store various types of data, including various data related to three-dimensional large intestine segmentation results, multiple connected components, midlines and endpoints, algorithm data, intermediate results, and various programs required to operate the device 500. The ROM 513 may be configured to store data and instructions for power-on self-test of the device 500, initialization of each functional module in the system, basic input / output driver programs of the system, and data and instructions required to boot the operating system.

[0037] Optionally, the device 500 may further include other hardware platforms or components, such as the shown tensor processing unit ("TPU") 514, graphics processing unit ("GPU") 515, field-programmable gate array ("FPGA") 516, and machine learning unit ("MLU") 517. It can be understood that although various hardware platforms or components are shown in the device 500, these are merely exemplary rather than restrictive, and those skilled in the art can add or remove corresponding hardware according to actual needs. For example, the device 500 may only include a CPU, related storage devices, and interface devices to implement the method for optimizing three-dimensional large intestine segmentation of the present application.

[0038] In some embodiments, for the convenience of data transfer and interaction with an external network, the device 500 of the present application further includes a communication interface 518, so that it can be connected to a local area network / wireless local area network ("LAN / WLAN") 505 through the communication interface 518, and then can be connected to a local server 506 or connected to the Internet ("Internet") 507 through the LAN / WLAN. Alternatively or additionally, the device 500 of the present application may also be directly connected to the Internet or a cellular network based on wireless communication technology through the communication interface 518, such as based on the 3rd generation ("3G"), 4th generation ("4G"), or 5th generation ("5G") wireless communication technology. In some application scenarios, the device 500 of the present application may also access a server 508 and a database 509 of an external network as needed, so as to obtain various known algorithms, data, and modules, and may remotely store various data, such as various data or instructions for presenting three-dimensional large intestine segmentation results, multiple connected components, midlines and endpoints, etc.

[0039] The peripheral devices of device 500 may include a display device 502, an input device 503, and a data transmission interface 504. In one embodiment, the display device 502 may include, for example, one or more speakers and / or one or more visual displays configured to provide voice prompts and / or image / video displays for the three-dimensional large intestine segmentation of the present application. The input device 503 may include, for example, a keyboard, a mouse, a microphone, a gesture capture camera, and other input buttons or controls configured to receive input of audio data and / or user instructions. The data transmission interface 504 may include, for example, a serial interface, a parallel interface, or a Universal Serial Bus interface ("USB"), a Small Computer System Interface ("SCSI"), Serial ATA, FireWire, PCI Express, and a High-Definition Multimedia Interface ("HDMI"), etc., configured for data transmission and interaction with other devices or systems. According to the solution of the present application, the data transmission interface 504 may receive the large intestine segmentation result segmented from the three-dimensional intestinal image collected by the CT device and transmit to the device 500 data or results including the large intestine segmentation result or various other types.

[0040] The above CPU 511, mass storage 512, ROM 513, TPU 514, GPU 515, FPGA 516, MLU 517, and communication interface 518 of the device 500 of the present application may be interconnected with each other through a bus 519 and achieve data interaction with the peripheral devices through this bus. In one embodiment, through this bus 519, the CPU 511 may control other hardware components and their peripheral devices in the device 500.

[0041] The above combination Figure 5 has described the device that can be used to execute the three-dimensional large intestine segmentation optimization of the present application. It should be understood that the device structure or architecture here is only exemplary, and the implementation manner and implementation entity of the present application are not limited by it, but can be changed without departing from the spirit of the present application.

[0042] According to the above description in conjunction with the drawings, those skilled in the art can also understand that the embodiments of the present application can also be implemented by a software program. Thus, the present application also provides a computer-readable storage medium on which computer-readable instructions for optimizing three-dimensional large intestine segmentation are stored. When the computer-readable instructions are executed by one or more processors, they can be used to implement the method for optimizing three-dimensional large intestine segmentation described in the present application in conjunction with the attached Figure 1 drawings.

[0043] It should be noted that although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the order of the steps depicted in the flowchart can be changed. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step for execution, and / or one step may be decomposed into multiple steps for execution.

[0044] It should be understood that when terms such as "first", "second", "third", and "fourth" are used in the claims, the specification, and the drawings of the present application, they are only used to distinguish different objects and not to describe a specific order. The terms "comprising" and "including" used in the specification and claims of the present application indicate the existence of the described features, wholes, steps, operations, elements, and / or components, but do not exclude the existence or addition of one or more other features, wholes, steps, operations, elements, components, and / or their combinations.

[0045] It should also be understood that the terms used in the specification of the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. As used in the specification and claims of the present application, unless the context clearly indicates otherwise, the singular forms "a", "an", and "the" are intended to include the plural forms. It should be further understood that the term "and / or" used in the specification and claims of the present application refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0046] Although the embodiments of the present application are as above, the above content is only an example used to facilitate the understanding of the present application and is not intended to limit the scope and application scenarios of the present application. Any person skilled in the technical field of the present application can make any modifications and changes in the form of implementation and details without departing from the spirit and scope disclosed by the present application. However, the scope of patent protection of the present application shall still be subject to the scope defined by the appended claims.

[0047] In addition, the collection and acquisition of various data in the present application comply with relevant laws and regulations and are authorized by the data provider. Any organization or individual that needs to obtain external data shall obtain authorization in accordance with the law and ensure data security, and shall not illegally collect, use, process, or transmit unauthorized or unprotected data, nor illegally buy, provide, or disclose unauthorized or unprotected data.

Claims

1. A method for optimizing three-dimensional large intestine segmentation, comprising: Obtaining a three-dimensional large intestine segmentation result, wherein the three-dimensional large intestine segmentation result contains a plurality of connected components; Extracting the endpoints of the centerlines of each of the connected components, and constructing an objective distance function based on the endpoints of the centerlines of each of the connected components; And Minimizing the objective distance function under the objective constraint conditions set based on 0-1 integer programming, such that the plurality of connected components form a single connected component, so as to optimize the three-dimensional large intestine segmentation result.

2. The method according to claim 1, wherein constructing an objective distance function based on the endpoints of the centerlines of each of the connected components comprises: Setting objective variables based on the endpoints of the centerlines of each of the connected components; And Constructing the objective distance function based on the product of the objective variables and the distances between the objective variables and the endpoints of the centerlines of each of the connected components.

3. The method according to claim 2, wherein the target distance function is constructed by the following formula: , wherein represents the distance between the endpoints of the center line of each connected component, represents the target variable, and i, j represent the endpoints of the center line of the connected component.

4. The method according to claim 2, wherein minimizing the objective distance function under the objective constraint conditions set based on 0-1 integer programming, such that the plurality of connected components form a single connected component, so as to optimize the three-dimensional large intestine segmentation result comprises: Determining objective endpoints by minimizing the objective distance function under the objective constraint conditions set based on 0-1 integer programming; And Connecting all the objective endpoints such that the plurality of connected components form a single connected component, so as to optimize the three-dimensional large intestine segmentation result.

5. The method according to claim 4, wherein the objective constraint conditions set based on 0-1 integer programming comprise: Setting the objective variables corresponding to the objective distances to be summed in the objective distance function to a first objective value, and setting the objective variables corresponding to the remaining objective distances to a second objective value; Restricting the values of the objective variables such that each of the connected components is connected to only one straight line segment; and Restricting the values of the objective endpoints to avoid adding redundant straight line segments.

6. The method according to claim 5, wherein the first objective value is 1 and the second objective value is 0.

7. The method according to claim 5, wherein the value of the target variable is restricted by the following formula so that each of the connected components is connected to only one straight line segment: and , where means that only one straight line is connected to the end point j, means that only one straight line is connected to the end point i.

8. The method according to claim 5, wherein the value of the target end point is restricted by the following formula to avoid adding redundant straight line segments: , wherein represents a straight line segment connecting the starting point and the ending point, represents an existing median line.

9. An apparatus for optimizing three-dimensional large intestine segmentation, comprising: A processor; And A memory storing computer instructions for optimizing three-dimensional large intestine segmentation, which when executed by the processor, cause the method according to any one of claims 1-8 to be implemented.

10. A computer-readable storage medium storing computer program instructions for optimizing three-dimensional large intestine segmentation, which when executed by one or more processors, cause the method according to any one of claims 1-8 to be implemented.

Images