Three-dimensional marker region determination method for target object and related device

Abstract

Embodiments of the present application provide a method for determining a three-dimensional marking area of a target object and related apparatus. The method comprises: obtaining a contour set of the target object, the set comprising object contours of the target object on a plurality of first sections of a three-dimensional image; determining a plurality of contour point groups based on the contour set, each contour point group comprising a single contour point on each object contour of the contour set, for a first contour point on a current first section and a second contour point on an adjacent first section adjacent to the current first section in each contour point group, the first contour point being the contour point closest to the second contour point in a target distance in a first target direction parallel to the plurality of first sections among the contour points on the current first section; generating an envelope line based on each contour point group in the plurality of contour point groups; and determining a three-dimensional marking area containing the target object based on the contour set and the generated envelope line. The scheme can improve the accuracy and efficiency of determining the three-dimensional marking area.

Description

Technical Field

[0001] This invention relates to the field of three-dimensional image processing technology, specifically to a method and apparatus for determining the three-dimensional marked region of a target object, an electronic device, a storage medium, and a computer program product. Background Technology

[0002] Currently, the processing of 3D images is widely used in various technical fields. To facilitate the analysis of target objects in 3D images, it is usually necessary to identify the location of the target object and obtain its 3D marked region. For example, in ultrasound imaging diagnosis, doctors often need to use ultrasound equipment to examine and diagnose target objects inside a patient's body (such as lesions in specific areas). In this process, the 3D marked region of the target object can be determined first, and the target object can be analyzed based on the 3D marked region, such as measuring specific parameters (e.g., area, diameter).

[0003] In related technologies, there are schemes that directly perform 3D region segmentation on 3D images or draw 2D contours to determine 3D marked regions. Schemes that directly perform 3D region segmentation on 3D images are algorithmically complex and time-consuming. Schemes that draw 2D contours have lower accuracy in identifying target object boundaries, poor edge smoothness, and the identified 3D marked regions are not accurate enough, easily leading to morphological errors. Furthermore, schemes that draw 2D contours are also time-consuming, requiring a significant workload for users. Summary of the Invention

[0004] The present invention addresses the aforementioned problems. It provides a method, apparatus, electronic device, and storage medium for determining a three-dimensional marked region of a target object. This solution effectively improves the accuracy and efficiency of determining the three-dimensional marked region, reduces user workload, and enhances user experience.

[0005] According to one aspect of the present invention, a method for determining a three-dimensional marked region of a target object is provided. The method includes: an acquisition step: acquiring a contour set of the target object, the contour set including object contours of the target object on multiple first cross-sections of a three-dimensional image, the multiple first cross-sections being cross-sections parallel to each other; a first determination step: determining multiple contour point groups based on the contour set, wherein each contour point group includes a single contour point on each object contour in the contour set, for a first contour point on a current first cross-section and a second contour point on an adjacent first cross-section adjacent to the current first cross-section in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first cross-section, the target distance being a distance in a first target direction, the first target direction being a direction parallel to the multiple first cross-sections; a generation step: generating an envelope based on each contour point group in the multiple contour point groups; and a second determination step: determining a three-dimensional marked region containing the target object based on the contour set and the generated envelope.

[0006] For example, generating an envelope based on each of a plurality of contour point groups includes: performing spline curve fitting based on each of the plurality of contour point groups to obtain an envelope corresponding to each contour point group.

[0007] For example, obtaining the contour set of the target object includes: for any one of a plurality of first cross-sections, determining the initial contour corresponding to the first cross-section; wherein the initial contour is the object contour; or, obtaining the contour set of the target object further includes: performing a preset additional processing on the initial contour corresponding to the first cross-section to determine the object contour corresponding to the first cross-section; wherein determining the initial contour corresponding to the first cross-section includes: performing automatic contour detection on the cross-sectional image corresponding to the first cross-section to obtain the initial contour corresponding to the first cross-section; or, obtaining object position information marked on the cross-sectional image corresponding to the first cross-section, determining the cross-sectional object region where the target object is located based on the pixel information at the position indicated by the object position information, and determining the contour of the cross-sectional object region as the initial contour corresponding to the first cross-section; or, obtaining a plurality of marked contour points marked on the cross-sectional image corresponding to the first cross-section, and determining the initial contour corresponding to the first cross-section based on the plurality of marked contour points; wherein the cross-sectional image is determined based on the corresponding first cross-section and the three-dimensional image.

[0008] For example, determining the initial contour corresponding to the first cross section based on multiple labeled contour points includes: upsampling or downsampling the multiple labeled contour points to obtain a new set of contour points; and performing spline curve fitting based on the new set of contour points to obtain the initial contour corresponding to the first cross section.

[0009] For example, the preset additional processing includes: adjusting the initial contour corresponding to the first cross-section in response to the input contour adjustment information to obtain the object contour corresponding to the first cross-section.

[0010] For example, determining the three-dimensional marker region containing the target object based on the contour set and the generated envelope includes: determining the three-dimensional object region where the target object is located based on the contour set and the generated envelope; expanding outward with a preset thickness based on the surface voxel points of the three-dimensional object region to obtain the expanded region as the three-dimensional marker region.

[0011] For example, the method further includes projecting the three-dimensional marker region and the three-dimensional object region onto the viewing surface for display.

[0012] For example, determining the 3D object region where the target object is located based on the contour set and the generated envelope includes: determining the value range of the target object on each coordinate axis in a preset 3D coordinate system according to the contour set and the envelope; establishing a cuboid-enclosed region of the target object based on the value range, wherein the length, width, and height of the cuboid-enclosed region are parallel to the three coordinate axes of the preset 3D coordinate system; for each of the multiple second sectional surfaces parallel to the preset coordinate axes in the preset 3D coordinate system, determining the sequence of intersection points between the sectional region of the cuboid-enclosed region on the second sectional surface and the envelope; constructing a closed polygon region based on the sequence of intersection points; assigning the voxel points contained in the closed polygon region to the first voxel value, and assigning the voxel points in the sectional region outside the closed polygon region to the second voxel value; determining the region where the voxel points in the cuboid-enclosed region have voxel values ​​equal to the first voxel values ​​as the 3D object region; wherein the preset coordinate axes are not perpendicular to the multiple first sectional surfaces, the spacing between the multiple second sectional surfaces in the second target direction is equal to the spacing between the voxel points in the cuboid-enclosed region in the second target direction, and the second target direction is a direction perpendicular to the multiple second sectional surfaces.

[0013] For example, determining the three-dimensional marked region containing the target object based on the contour set and the generated envelope further includes: calculating the gradient value of each voxel point within the cuboid-enclosed region; and determining voxel points with gradient values ​​greater than a preset gradient threshold as surface voxel points of the three-dimensional object region.

[0014] Exemplarily, the method further includes: updating the contour set to obtain a new contour set, and re-executing the first determining step, the generating step, and the second determining step for the new contour set; wherein the updating operation includes one or more of the following operations: adding a new object contour to the contour set in response to an addition operation to an object contour in the contour set; deleting at least a portion of the object contours from the contour set in response to a deletion operation to an object contour in the contour set; generating one or more new object contours in response to a contour adjustment operation to one or more object contours in the contour set, and replacing the corresponding object contours in the contour set with the newly generated object contours to update the contour set.

[0015] For example, determining multiple contour point groups based on a contour set includes: performing smooth resampling processing on multiple object contours in the contour set to obtain a contour point set uniquely corresponding to each object contour; determining multiple contour point groups based on the contour point sets corresponding to multiple object contours; wherein all contour points in each contour point set belong to different contour point groups in multiple contour point groups.

[0016] According to another aspect of the present invention, a three-dimensional marking region determination apparatus for a target object is provided. The apparatus includes: an acquisition module for acquiring a contour set of the target object, the contour set including object contours of the target object on multiple first cross-sections of a three-dimensional image, the multiple first cross-sections being cross-sections parallel to each other; a first determination module for determining multiple contour point groups based on the contour set, wherein each contour point group includes a single contour point on each object contour in the contour set, for a first contour point on a current first cross-section and a second contour point on an adjacent first cross-section adjacent to the current first cross-section in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first cross-section, the target distance being a distance in a first target direction, the first target direction being a direction parallel to the multiple first cross-sections; a generation module for generating an envelope based on each contour point group in the multiple contour point groups; and a second determination module for determining a three-dimensional marking region containing the target object based on the contour set and the generated envelope.

[0017] According to another aspect of the present invention, an electronic device is provided, including a processor and a memory, wherein the memory stores computer program instructions, which are executed by the processor to perform the above-described method for determining the three-dimensional marked region of the target object.

[0018] According to another aspect of the present invention, a storage medium is provided that stores a computer program / instructions, which, when executed, are used to perform the above-described method for determining the three-dimensional marked region of the target object.

[0019] According to another aspect of the present invention, a computer program product is provided, including computer program instructions, which, when executed, are used to perform the above-described method for determining the three-dimensional marked region of the target object.

[0020] According to embodiments of the present invention, a method, apparatus, electronic device, storage medium, and computer program product for determining a three-dimensional marked region of a target object can acquire the object contour of the target object on multiple first cross-sections of a three-dimensional image, determine contour point groups based on each object contour, and determine the three-dimensional marked region of the target object using the envelope generated based on the contour point groups and the object contour. In this method, the contour points in each contour point group are obtained based on nearest neighbor search, which ensures that the envelope does not intersect, thereby effectively avoiding geometric errors when determining the three-dimensional marked region. Therefore, the three-dimensional marked region determined by this scheme has high accuracy and stability. Furthermore, this scheme can automatically generate the envelope based on the object contour without manual intervention, and the algorithm for generating the envelope is relatively simple, allowing for rapid determination of the three-dimensional marked region of the target object. In summary, this scheme can effectively improve the accuracy and efficiency of determining the three-dimensional marked region, reduce user workload, and enhance user experience.

[0021] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0022] The above and other objects, features, and advantages of the present invention will become more apparent from the more detailed description of the embodiments of the invention in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same parts or steps.

[0023] Figure 1 A schematic flowchart of a method for determining a three-dimensional marked region of a target object according to an embodiment of the present invention is shown;

[0024] Figure 2 A schematic diagram of a plurality of first cross-sections according to an embodiment of the present invention is shown;

[0025] Figure 3 A schematic diagram illustrating the determination of a set of contour points according to an embodiment of the present invention is shown;

[0026] Figure 4A schematic diagram of marked contour points on a first cross-section according to an embodiment of the present invention is shown;

[0027] Figure 5 A rendering of a three-dimensional marked region of a lesion according to an embodiment of the present invention is shown;

[0028] Figure 6a , 6b 6c and 6d respectively show schematic diagrams of lesions on different sections according to an embodiment of the present invention;

[0029] Figure 7 A schematic flowchart illustrating a method for determining a three-dimensional marked region of a target object according to an embodiment of the present invention is shown.

[0030] Figure 8 A schematic block diagram of a three-dimensional marking region determination device for a target object according to an embodiment of the present invention is shown; and

[0031] Figure 9 A schematic block diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of the present invention more apparent, exemplary embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments described in the present invention, all other embodiments obtained by those skilled in the art without inventive effort should fall within the protection scope of the present invention.

[0033] As mentioned above, to facilitate the analysis of target objects in 3D images, it is usually necessary to identify the location of the target object and obtain its 3D marked region. The following explanation uses the marking of lesions in the medical field as an example. Lesion marking refers to the indication of lesion regions on medical images in the form of contour markings or global markings. The global marking can be a region mask of the lesion region or the image pixel area occupied by the lesion region in the 3D image. Those skilled in the art will understand that a region mask of the lesion region refers to a marking method that distinguishes pixels at the location of the lesion region from pixels at other locations using different pixel values ​​(e.g., 1 and 0). Lesion markings can be divided into two-dimensional markings and three-dimensional markings based on image dimensional characteristics. Three-dimensional markings are the division of three-dimensional lesion regions, containing more comprehensive information about the lesion's extent. After obtaining the three-dimensional markings, two-dimensional markings (two-dimensional contour markings or two-dimensional global markings) can be obtained through methods of regenerating cross-sections in any orientation. For example, in the fusion imaging function of ultrasound equipment, tumor lesions are typically located and 3D-labeled based on volumetric images (i.e., three-dimensional images) such as second-modality computed tomography (CT) images / magnetic resonance (MR) images. These 3D labels can be displayed directly in a 3D viewport, or their corresponding 2D labels can be displayed on standard or free-viewing planes. Through image registration and fusion processes, 3D labels on CT or MR images can be mapped onto ultrasound 2D planes, enabling real-time reference of target object markings during ultrasound examinations.

[0034] From an implementation perspective, the methods for 3D marking of target objects on 3D images in related technologies can be divided into two categories: (1) using artificial intelligence (AI) and other methods to perform automated 3D region segmentation and directly determine the position of the 3D marking region; (2) responding to input operations such as clicking performed on the 3D image through an input device (e.g., the touchpad of an ultrasound device), drawing the 2D contour of the target object and then generating the 3D marking region. Both methods have their advantages and disadvantages. The first method is theoretically better, as there is no deviation caused by manual intervention, but the algorithm used in this method is more complex and usually takes longer. The second method draws 2D contours on multiple parallel and vertical sections through an input device and then generates the 3D marking region. The problems with the second method include: the contours on different sections generated are not accurately aligned, there are topological problems, and it is easy to cause errors in the section shape. Therefore, the accuracy of the 3D marking region generated by this scheme is not high enough and the anti-interference ability is insufficient. In addition, this scheme takes a long time (on the order of 10 seconds) due to being purely manual, and the user's workload is large.

[0035] To at least partially solve the above-mentioned problems, embodiments of the present invention provide a method for determining a three-dimensional marked region of a target object. The present invention establishes a novel three-dimensional marked region determination method based on cross-sectional contours. This method can be used for applications such as three-dimensional lesion marking on medical body images, and has advantages such as accurate marking results, stable generation method, and short overall processing time. Using the three-dimensional marked region determination method of the embodiments of the present invention, three-dimensional marked regions can be conveniently drawn on three-dimensional images such as CT / MR. Exemplarily, the method for determining the three-dimensional marked region of a target object can run on any electronic device. This electronic device can be any device with data processing capabilities and / or instruction execution capabilities. The electronic device can include, but is not limited to, one or more of personal computers, servers, mobile terminals, and ultrasound equipment. When the electronic device is independent of the ultrasound equipment, the electronic device can optionally be communicatively connected to the ultrasound equipment to acquire three-dimensional images from the ultrasound equipment. Of course, the three-dimensional images acquired by the ultrasound equipment can also be transferred from the ultrasound equipment to the aforementioned electronic device by copying using a storage device such as flash memory. Exemplarily, the electronic device for executing the method for determining the three-dimensional marked region of a target object may include a processing device. In the electronic device, the method for determining the three-dimensional marked region of the target object can be specifically executed by the processing device. Exemplarily, an electronic device for performing a method for determining a three-dimensional marked region of a target object may further include a display device and / or an input device, or be communicatively connected to a display device and / or an input device. The input device may include, but is not limited to, one or more of the following: a keyboard, a mouse, a touchscreen, a touchpad, etc. A touchpad may be referred to as a control panel. The input device may be, for example, a touchpad including a trackball and multiple function keys. The display device may be used to display one or more of the three-dimensional image, three-dimensional marked region, first region outline, and second region outline described herein.

[0036] Figure 1 A schematic flowchart illustrating a method for determining a three-dimensional marked region of a target object according to an embodiment of the present invention is shown. Figure 1 As shown, the method for determining the three-dimensional marked region of the target object includes an acquisition step S110, a first determination step S120, a generation step S130, and a second determination step S140.

[0037] Step S110: Obtain the contour set of the target object. The contour set includes the object contours of the target object on multiple first cross-sections of the 3D image. The multiple first cross-sections are cross-sections that are parallel to each other.

[0038] The target object can represent any specific part or tissue of a living organism such as a human or animal, or a specific region (e.g., a lesion area) within any specific part or tissue. The 3D image can be any 3D image containing the target object. For example, the 3D image can be such as a CT image, MR image, or ultrasound image. The contour set includes multiple object contours, each of which can include a sequence of discrete contour points. For example, and not limitingly, an object contour can be represented by the sequence of contour points it contains. Obtaining an object contour can be obtaining the sequence of contour points on the object contour. The object contour is the cross-sectional contour of the target object on a first cross-section. The first cross-section can be any cross-section, including but not limited to transverse, coronal, sagittal, or cross-sections perpendicular to the target viewing direction. The target viewing direction is the direction the user currently expects to observe; it can be the view direction of the currently displayed 3D ultrasound image on the screen. It is understood that the target viewing direction can change when the user adjusts the viewing angle of the 3D ultrasound image of the target object. Figure 2 A schematic diagram of a plurality of first cross-sections according to an embodiment of the present invention is shown. Figure 2 As shown, the acquired 3D image of the target object can be imported into the visualization software of an electronic device. The visualization software can be run by the processing unit of the electronic device, which then controls the display device to display the image on its interface. Figure 2 The visualization software window shown is illustrated. For example, the 3D image of the target object can be segmented along multiple mutually parallel first facets perpendicular to the current visualization software window direction to obtain a facet image corresponding to each first facet. For each facet image, the contour of the target object on the facet image can be determined by any method, such as automatically recognizing the contour using an algorithm or determining the contour in response to input operations such as clicks on an input device, thus obtaining the object contour. In this document, for ease of description, the multiple first facets are ordered and labeled along directions perpendicular to the first facets, for example, each first facet is labeled as layer 1, layer 2, layer 3... layer N (e.g., layer N). Figure 2 As shown in the figure, N is the total number of layers in the first cut surface.

[0039] The number of first facets and the spacing between adjacent first facets can be set as needed. The number of first facets and / or the spacing between adjacent first facets can be preset to a fixed value, or can be flexibly set according to actual needs. It is understood that the denser the first facets, the more accurate the three-dimensional marking area of ​​the target object will be subsequently determined. Exemplarily, but not limitingly, the number of first facets can be in the range of, for example, [3, 10]. For example, 5, 6, 8, etc. Multiple first facets can optionally be set to a uniform interval. On the display interface, the three-dimensional image can be displayed with multiple facets at any angle. Exemplarily, a suitable facet can be selected as the required first facet as needed, and the object outline can be marked on each first facet. For example, in response to the display command received by the input device, the cross section, coronal plane, sagittal plane, or any free facet with adjusted azimuth angle of the three-dimensional image can be displayed on the display interface for the user to observe and select the facet containing the most layers of the target object for layered outline drawing. For example, in response to operations on the layer shift settings control on the display interface, the current viewport can be shifted, and in response to operations on the selection control on the display interface, facets that intersect with the target object can be selected at regular intervals along a direction perpendicular to the viewport. The number of facets should ideally be 3-6 layers, and the intervals should ideally be uniform. For example, if using the same facet interval yields the most layers when using the coronal plane as the first facet, then the coronal plane can be selected as the first facet. Furthermore, when selecting multiple first facets, the selected facets should ideally include the beginning and end positions of the target object; that is, the selected facets should ideally include the first facets corresponding to the positions where the target object first appears and disappears in the vertical direction.

[0040] First determination step S120: Determine multiple contour point groups based on the contour set, wherein each contour point group includes a single contour point on each object contour in the contour set. For a first contour point on the current first sectional plane and a second contour point on an adjacent first sectional plane adjacent to the current first sectional plane in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first sectional plane. The target distance is the distance in the first target direction, which is a direction parallel to the multiple first sectional planes.

[0041] A contour point can be obtained from each object contour in the contour set, and the contour points obtained from different object contours can form a contour point group. By changing the position of the contour points obtained from each object contour, another contour point group can be obtained using the above method. Thus, multiple contour point groups can be obtained as needed. The number of contour point groups can be any number greater than or equal to 2. For each contour point group, the contour points within it conform to the following rule: for a first contour point on the current first sectional plane and a second contour point on an adjacent first sectional plane, the first contour point is the contour point on the current first sectional plane with the closest target distance to the second contour point. The contour points in a contour point group only need to conform to the above rule; the method for determining the contour point group is arbitrary. For example, one can first select any contour point (or selected contour point) on any first cross-section (or first selected cross-section); then, from the first cross-section adjacent to the first selected cross-section (or second selected cross-section), select the contour point with the closest target distance to the selected contour point on the first selected cross-section, thus obtaining the selected contour point on the second selected cross-section; then, from the first cross-section adjacent to the second selected cross-section (or third selected cross-section), select the contour point with the closest target distance to the selected contour point on the second selected cross-section, thus obtaining the selected contour point on the third selected cross-section; then, from the first cross-section adjacent to the third selected cross-section (or fourth selected cross-section), select the contour point with the closest target distance to the selected contour point on the third selected cross-section, thus obtaining the selected contour point on the fourth selected cross-section; and so on, until all contour points on the first cross-sections have been selected. Finally, the selected contour points selected from each first cross-section can form a contour point group. The first selected section can be any one of multiple first sections. It can be any one of the outermost first sections or any one of the first sections that is not located on the outermost side.

[0042] To facilitate understanding, the relationship between the first contour point and the second contour point is illustrated below with an example. Assume that the first cross-section is represented as a horizontal (i.e., transverse) cross-section in the current viewing view of the presented 3D ultrasound image, meaning the first cross-section is parallel to the horizontal direction. Then, the first target direction can be the horizontal direction, and the target distance can refer to the horizontal distance between the two contour points (which can be called the horizontal distance). In other words, the first contour point is the contour point on the current first cross-section that has the closest horizontal distance to the second contour point. The distance between the two contour points in the first target direction (i.e., the target distance) can be calculated using any distance measurement method, including but not limited to the method described below, which involves projecting one contour point onto the first cross-section where the other contour point is located and then calculating the distance between the projected point and the other contour point. An exemplary method for determining a group of contour points is described below using a distance measurement method based on projection points.

[0043] For example, multiple first cut surfaces can be ordered along a direction perpendicular to the first target direction (i.e., a direction perpendicular to the first cut surface). The adjacent first cut surfaces can be either the previous first cut surface located above the current first cut surface in sequence, or the next first cut surface located below the current first cut surface in sequence. It is understood that when the first cut surface is the aforementioned horizontal cut surface, multiple first cut surfaces can be ordered along a vertical direction. For example, the first contour point can be the contour point on the current first cut surface that is closest to the projection point of the second contour point. The projection point is the projection point obtained by projecting the second contour point onto the current first cut surface along a direction perpendicular to the first target direction. For example, the first cut surfaces can be ordered along a direction perpendicular to them, identified as layer 1, layer 2, layer 3... layer N as described above, where N is the total number of layers of the first cut surface. Multiple contour points (i.e., the aforementioned sequence of contour points) can be determined on the object contour of each first cut surface, and contour points used to generate the longitudinal connection curve (i.e., the envelope) can be sequentially searched on each layer of the first cut surface. For example, starting from a contour point on the first layer (which can be called the j-th contour point), the first sectional plane of the next layer can be determined along the vertical direction, and the corresponding contour point can be searched on the object contour of the first sectional plane of the next layer. The nearest neighbor contour point to the projection point of the j-th contour point can be automatically selected based on distance sorting and added to the contour point group corresponding to the j-th contour point. The projection point of the j-th contour point is the projection point obtained by projecting the j-th contour point onto the current first sectional plane along a direction perpendicular to the first sectional plane. j = 1, 2, 3, …, M, where M is the total number of contour points on the object contour of the first layer. Figure 3 A schematic diagram illustrating the determination of a set of contour points according to an embodiment of the present invention is shown. Figure 3As shown, the coordinates of the contour points on the first sectional surface of the first layer are (1,1), (1,2), (1,3), and (1,4). In the direction perpendicular to the first sectional surface, the contour point closest to the projection point of (1,1) on the first sectional surface of the second layer is contour point (2,1). Therefore, contour points (1,1) and (2,1) can be added to the first contour point group. Similarly, (1,2) and (2,2) can be added to the second contour point group, (1,3) and (2,3) to the third contour point group, (1,4) and (2,4) to the fourth contour point group, and so on. If the first sectional surface of the first layer contains M contour points, then M contour point groups can be formed. For the j-th contour point on the first sectional surface of the k-th layer, the nearest neighbor contour point on the (k+1)-th layer of the first sectional surface that is closest to the projection point of the contour point can be determined by the nearest neighbor search method described above, and the determined contour point is added to the contour point group corresponding to the j-th contour point on the first sectional surface. k = 1, 2, 3, …, N, where N is the total number of layers of the first sectional surface.

[0044] The number of layers in the first facet can be set to be relatively small, allowing the steps for determining the contour point group described above to yield results more quickly. Using a nearest neighbor search strategy ensures that vertically connected curves do not intersect, effectively preventing geometric errors when subsequently determining the 3D marked area.

[0045] Generation step S130: Generate an envelope based on each of the multiple contour point groups.

[0046] For each contour point group, curve fitting can be performed based on the contour points in that group to form the corresponding envelope. Since the contour points in each group are obtained using a nearest neighbor search, there are no intersections between the vertically connected envelopes. This hierarchical curve fitting method has low computational complexity and low storage overhead, taking only 10-100 milliseconds (ms) under normal conditions, resulting in fast response times for both initial envelope establishment and subsequent envelope adjustments. Furthermore, the inter-layer curve fitting connection method ensures the smoothness of the 3D marked region, leading to high accuracy.

[0047] The second determination step S140: Determine the three-dimensional marked region containing the target object based on the contour set and the generated envelope.

[0048] A three-dimensional marked region can be used to indicate the position of a target object in three-dimensional space. The object contours on multiple first cross-sections in the horizontal direction and the envelope formed by each set of contour points in the vertical direction can form an envelope volume, which can represent the three-dimensional contour of the target object. Therefore, the three-dimensional marked region of the target object can be determined through the set of contours and the generated envelope volume. The extent of the current three-dimensional marking can be initially viewed through the envelope volume. Exemplarily, and not limitingly, after confirming the envelope volume, marking information for the three-dimensional marked region of the target object can be further generated. The marking information for the three-dimensional marked region can include the contour markings of the three-dimensional marked region and / or the global markings of the three-dimensional marked region. As described above, the global markings of the three-dimensional marked region can be a region mask of the three-dimensional marked region or the image pixel area occupied by the three-dimensional marked region in the three-dimensional image.

[0049] In the above embodiments, the object contours of the target object on multiple first cross-sections of a 3D image can be obtained. Contour point groups are determined based on each object contour, and the 3D marked region of the target object is determined using the envelope generated based on the contour point groups and the object contours. In this method, the contour points in each contour point group are obtained based on nearest neighbor search, which ensures that the envelopes do not intersect, thus effectively avoiding geometric errors when determining the 3D marked region. Therefore, the 3D marked region determined by this scheme has high accuracy and stability. Furthermore, this scheme can automatically generate envelopes based on the object contours without manual intervention, and the algorithm for generating envelopes is relatively simple, allowing for rapid determination of the 3D marked region of the target object. In summary, this scheme can effectively improve the accuracy and efficiency of determining the 3D marked region, reduce user workload, and enhance user experience.

[0050] In some embodiments, generating an envelope based on each of the multiple contour point groups (i.e., generation step S130) includes: performing spline curve fitting based on each of the multiple contour point groups to obtain an envelope corresponding to each contour point group.

[0051] For example, for each contour point group, control points used by the spline curve can be determined based on the contour points in the contour point group, and spline curve fitting can be performed based on the control points to form the corresponding envelope.

[0052] The basic formula for spline curves is:

[0053] ; Formula (1)

[0054] in, Indicates the first The coordinates of the control points , Indicates the first The coordinates of the shape value points are given, where n represents the total number of shape value points. In this embodiment, the shape value points are the contour points in each contour point group, i.e., n = N. Solving the system of equations (1) simultaneously for a total of n gives the control point set. The coordinates of any point on a segment of the fitted spline curve can be calculated using the four nearest neighbors in the control point set, as follows:

[0055] ; Formula (2)

[0056] Here, the four nearest neighbors are used to determine the curve segment. Their respective coefficients, This is the path coefficient, with values ​​ranging from 0 to 1. Depend on The answer depends on the specific interpolation method chosen. Existing interpolation methods such as linear interpolation can be used. Linear interpolation can be used to estimate the value of an unknown point between known data points.

[0057] Exemplary and not restrictive, The specific formula for determining this can be as follows:

[0058] ; Formula (3)

[0059] Using the above method, the coordinates of any point on the fitted spline curve (i.e., envelope) corresponding to each contour point group can be calculated, thus determining the envelope corresponding to each contour point group. Of course, the above method is merely an example; other suitable existing or future curve fitting methods can be used to generate the envelope. For example, one or more of the following curve fitting methods can be used to generate the envelope: least squares curve fitting algorithm, cubic natural spline curve fitting algorithm, Hermite spline curve fitting algorithm, cardinal spline curve fitting algorithm, Bezier spline curve fitting algorithm, etc.

[0060] In the above embodiments, spline curve fitting is used to determine the corresponding envelope for each set of contour points. This envelope generation scheme is simple to implement and has a fast calculation speed, which can further improve the efficiency of determining the three-dimensional marking region. In addition, the gradient continuity of the spline curve makes the envelope surface sufficiently smooth, which helps to further improve the accuracy of the determined three-dimensional marking region.

[0061] In some embodiments, obtaining the contour set of the target object (i.e., obtaining step S110) includes: for any one of the plurality of first cut surfaces, determining the initial contour corresponding to the first cut surface; wherein the initial contour is the object contour, or, obtaining the contour set of the target object further includes: for any one of the plurality of first cut surfaces, performing a preset additional processing on the initial contour corresponding to the first cut surface to determine the object contour corresponding to the first cut surface.

[0062] The initial contour corresponding to the first cross-section can be determined through various methods. For example, after determining the initial contour corresponding to any first cross-section, this initial contour can be directly used as the object contour corresponding to the first cross-section. Alternatively, after determining the initial contour corresponding to any first cross-section, further processing (such as contour point interpolation, contour adjustment based on user input information, etc.) can be performed on the initial contour to obtain the object contour corresponding to the first cross-section. For any two different first cross-sections, both can use their corresponding initial contours as the corresponding object contours, or both can perform further processing on their corresponding initial contours to determine the corresponding object contours. Alternatively, one first cross-section can use its corresponding initial contour as the corresponding object contour, while the other first cross-section performs further processing on its corresponding initial contour to determine the corresponding object contour.

[0063] For example, determining the initial contour corresponding to the first cross-section may include: performing automatic contour detection on the cross-section image corresponding to the first cross-section to obtain the initial contour corresponding to the first cross-section; wherein the cross-section image is determined based on the three-dimensional image and the corresponding first cross-section.

[0064] The aforementioned automatic contour detection can be implemented using any existing or future contour detection algorithm. For example, a two-dimensional image segmentation model can be used to determine the initial contour of the first cross-section. The cross-section image corresponding to the first cross-section is input directly or after preprocessing into the two-dimensional image segmentation model, which can output the corresponding image segmentation result. Those skilled in the art will understand how to obtain the two-dimensional cross-section image corresponding to any cross-section from a three-dimensional image, which will not be elaborated here. Preprocessing operations may include, but are not limited to, resizing, normalization, and image noise reduction. The image segmentation result may include the contour position information of the target object's contour (i.e., the initial contour corresponding to the first cross-section) and / or the mask information of the target object in the cross-section image. When outputting mask information in the image segmentation result, the contour of the target object can be further determined based on the mask information. This automatic contour detection scheme can reduce the user's workload and provide a better user experience.

[0065] Optionally, the two-dimensional image segmentation model may include, but is not limited to, one or more of Fully Convolutional Networks (FCN), U-Net, and Mask Region-based Convolutional Neural Networks (Mak-RCNN). In the embodiments of this invention, no limitation is made to the two-dimensional image segmentation model. Any model capable of image segmentation is within the scope of protection of this invention.

[0066] For example, determining the initial contour corresponding to the first cross-section may include: obtaining object position information marked on the cross-section image corresponding to the first cross-section, determining the cross-section object region where the target object is located based on the pixel information at the position indicated by the object position information, and determining the contour of the cross-section object region as the initial contour corresponding to the first cross-section; wherein, the cross-section image is determined based on the three-dimensional image and the corresponding first cross-section.

[0067] For example, a user can click or drag to select any location on the cross-sectional image using an input device such as a mouse or trackball. In response to the user's selection, the location information of the user-selected location (which can be called the labeled location) can be obtained as the labeled object location information. The object location information is used to indicate the location of the target object. The number of labeled locations can be one or more, and each labeled location can be any location on the target object. For the cross-sectional image, the pixel values ​​of pixels at other locations in the image can be compared with the pixel values ​​at the labeled locations. For example, pixels whose pixel values ​​differ from the pixel values ​​at the labeled locations by a preset difference threshold can be identified as pixels belonging to the cross-sectional object region. The cross-sectional object region is the region where the target object is located; this region can be called the foreground region (which is a two-dimensional foreground region). Regions where no target object exists can be called the background region (which is a two-dimensional background region). Those skilled in the art will understand the meaning of foreground and background regions, and will not elaborate further here. The location of the cross-sectional object region in the cross-sectional image can be determined in the above manner. The contour of the acquired cross-sectional object region can be determined as the initial contour corresponding to the first cross-section. This approach can determine the initial contour based on the labeled object position information, which means it can draw on user experience to a certain extent to achieve contour detection of the target object, thus helping to improve the accuracy of contour determination in complex image scenes, for example.

[0068] For example, determining the initial contour corresponding to the first section may include: obtaining multiple annotation contour points marked on the section image corresponding to the first section, and determining the initial contour corresponding to the first section based on the multiple annotation contour points.

[0069] Similar to labeling locations, multiple contour points on the outline of a target object can be labeled by clicking on them using an input device such as a mouse or trackball. These labeled contour points are called labeled contour points. Figure 4 A schematic diagram showing the marked contour points on a first cross-section according to an embodiment of the present invention is shown. Figure 4 As shown, several (e.g., 4 to 8) contour points can be marked along the contour edge of the target object on the first cross-section. Using these multiple contour points marked on the first cross-section, curve fitting can be performed to fit the contour curve of the target object, thus obtaining the initial contour. The method of obtaining the initial contour based on the marked contour points is similar to the method described above for obtaining the envelope based on the contour points in the contour point group. The curve fitting method in this embodiment can be understood by referring to the curve fitting method described above, and will not be elaborated here. Optionally, curve fitting can be performed directly based on multiple marked contour points, or multiple marked contour points can be upsampled or downsampled before curve fitting is performed based on the contour points in the sampling results. This approach can determine the initial contour based on the marked contour points, and it can, to some extent, draw on user experience to achieve contour detection of the target object, helping to improve the accuracy of contour determination in complex image scenes, for example. Furthermore, this approach can reduce the dependence on image signal-to-noise ratio and the homogeneity of the target object, which also helps to improve the accuracy of contour determination.

[0070] In manual annotation schemes, users typically only need to input 20 to 30 annotation locations or points in the entire process of determining the 3D annotation area, thus reducing the user's workload.

[0071] In the above embodiments, different methods can be used to obtain the contour set of the target object in different actual scenarios, which helps to improve the accuracy and robustness of the method for determining the three-dimensional marked region of the target object, making the obtained contour set of the target object more reliable and practical.

[0072] In some embodiments, determining the initial contour corresponding to the first cross-section based on multiple labeled contour points includes: upsampling or downsampling the multiple labeled contour points to obtain a new set of contour points; and performing spline curve fitting based on the new set of contour points to obtain the initial contour corresponding to the first cross-section.

[0073] Based on the labeled contour points in each first cross-section, upsampling or downsampling can be performed to obtain a new set of contour points. Upsampling can increase the sampling rate through certain methods, thereby improving the accuracy and quality of the contour point data. Downsampling can extract the main features in the data, reduce redundant information, and reduce computational burden. Optionally, upsampling methods can include nearest neighbor interpolation, bilinear interpolation, cubic spline interpolation, deconvolution, etc. Downsampling can be implemented by methods such as extracting contour points at fixed intervals.

[0074] Based on the new set of contour points, spline curve fitting can be performed to determine the contour curve corresponding to each first cross-section, thus obtaining the initial contour. Spline curve fitting can be achieved using formulas (1), (2), and (3) as described above. When calculating formula (1), the shape points used are the contour points in the new set of contour points, and the total number of shape points n is equal to the total number of contour points in the new set of contour points.

[0075] In the above embodiments, a new set of contour points is obtained by upsampling or downsampling, which can better balance curve fitting accuracy and computational efficiency as needed, so as to ensure the accuracy of the three-dimensional marking area of ​​the target object while obtaining a high image processing speed.

[0076] In some embodiments, the preset additional processing includes: adjusting the initial contour corresponding to the first cross-section in response to the input contour adjustment information to obtain the object contour corresponding to the first cross-section.

[0077] After drawing the initial contour corresponding to the first cross-section, if the obtained initial contour deviates from the user's expectations, the user can use controls to move any contour point, a segment of the initial contour, or the entire initial contour to a suitable position, and recalculate and update the contour curve in real time to complete the adjustment of the initial contour. The adjusted initial contour can be used as the object contour. For example, when the user determines that the adjusted initial contour meets the expected requirements, they can input a confirmation command (which can be called the first confirmation command) through the mouse, trackball, keyboard, etc. Upon receiving the user's first confirmation command, the current initial contour can be designated as the object contour.

[0078] In the above embodiments, the initial contour corresponding to the first cross-section can be flexibly adjusted according to the actual situation, which can make the formed object contour fit the target object better, further improving the accuracy of the three-dimensional marked area result of the formed target object and the flexibility of the method.

[0079] In some embodiments, determining the three-dimensional marker region containing the target object based on the contour set and the generated envelope (second determination step S140) includes: determining the three-dimensional object region where the target object is located based on the contour set and the generated envelope; expanding outward with a preset thickness based on the surface voxel points of the three-dimensional object region to obtain the expanded region as the three-dimensional marker region.

[0080] Devices used for acquiring three-dimensional images (such as ultrasound equipment) can acquire images of a three-dimensional spatial region to obtain a three-dimensional image. The three-dimensional spatial region can include the three-dimensional object region containing the target object and spatial regions that do not contain the target object. The three-dimensional object region containing the target object can be called the foreground region (it is a three-dimensional foreground region). The spatial region that does not contain the target object can be called the background region (it is a three-dimensional background region). Optionally, the three-dimensional object region can be determined based on a set of contours and a generated envelope. Exemplarily, but not limitingly, a three-dimensional mask image can be generated based on the set of contours and the generated envelope to identify the location of the three-dimensional object region. For example, the three-dimensional mask image can be the same size as the originally acquired three-dimensional image, and voxel points corresponding to the target object can be marked with a first voxel value (e.g., 1) in the three-dimensional mask image, and voxel points corresponding to the background region can be marked with a second voxel value (e.g., 0). It is understood that the position of the voxel point marked with the first voxel value in the three-dimensional mask image can represent the position of the three-dimensional marked region in the three-dimensional image. The voxel point marked with the first voxel value can be regarded as a region mask of the three-dimensional marked region described above. After determining the location of the 3D object region, the 3D object region can be identified from the 3D image, and the surface voxel points of the 3D object region can be determined.

[0081] The 3D object region is expanded using surface voxel points as a reference, that is, the surface of the 3D object region is extended by a certain thickness. This creates an extended safe area, effectively preventing the omission of markers. For example, a specific bitwise operation kernel can be used to expand the shape of the foreground region in the 3D image or 3D mask image. In the 3D case, using a binary spherical kernel can avoid jagged edges. The 3D marker region is the expanded region, which can include both the 3D object region and the safe area. For example, to clearly display the safe area within the expanded 3D marker region, the voxel points corresponding to the safe area can be marked with a third voxel value (e.g., 2) different from the first and second voxel values ​​in the 3D mask image. The user can adjust the preset thickness according to the actual situation. For example, the preset thickness can range from [3, 7] millimeters (mm). For example, the preset thickness can be 3mm, 5mm, 7mm, etc., preferably 5mm.

[0082] It should be noted that the above expansion scheme is only an example. Optionally, the three-dimensional object region can be directly used as the three-dimensional marker region to simplify the operation process of the method for determining the three-dimensional marker region of the target object and improve the running efficiency.

[0083] In the above embodiments, the surface voxel points of the three-dimensional object region are used as a reference to expand outward with a preset thickness, and the expanded region is used as a three-dimensional marking region. This makes it convenient for users to take into account the information in the safe area when performing subsequent analysis or operation on the target object, thereby effectively reducing the adverse effects on the subsequent analysis or operation of the target object caused by inaccurate positioning of the three-dimensional object region.

[0084] In some embodiments, determining the 3D object region where the target object is located based on the contour set and the generated envelope includes: determining the value range of the target object on each coordinate axis in a preset 3D coordinate system according to the contour set and the envelope; establishing a cuboid-enclosed region of the target object based on the value range, wherein the length, width, and height of the cuboid-enclosed region are parallel to the three coordinate axes of the preset 3D coordinate system in a one-to-one correspondence; for each of the multiple second sectional surfaces parallel to the preset coordinate axes in the preset 3D coordinate system, determining the sequence of intersection points between the sectional region of the cuboid-enclosed region on the second sectional surface and the envelope; constructing a closed polygon region based on the sequence of intersection points; assigning the voxel points contained in the closed polygon region to a first voxel value, and assigning the voxel points in the sectional region located outside the closed polygon region to a second voxel value; determining the region where the voxel points in the cuboid-enclosed region have voxel values ​​equal to the first voxel values ​​as the 3D object region; wherein the preset coordinate axes are not perpendicular to the multiple first sectional surfaces, the spacing between the multiple second sectional surfaces in the second target direction is equal to the voxel point spacing of the cuboid-enclosed region in the second target direction, and the second target direction is a direction perpendicular to the multiple second sectional surfaces.

[0085] The preset 3D coordinate system can be, for example, a 3D Cartesian coordinate system. Exemplarily, the preset 3D coordinate system can be the 3D coordinate system used by the 3D image, or it can be a 3D coordinate system whose XYZ axes are parallel to the XYZ axes used by the 3D image but have different origins. Exemplarily, the origin of the preset 3D coordinate system can be selected at any position as needed. Of course, the preset 3D coordinate system can also be a 3D coordinate system with arbitrary XYZ axis orientations. The contour set and envelope can form the envelope of the target object. Placing the envelope into the preset 3D coordinate system determines the range of values ​​for the target object on the three coordinate axes of the preset 3D coordinate system: the X-axis, Y-axis, and Z-axis. A cuboid-shaped bounding box region (referred to herein as the cuboid bounding region) can be established using the above value range as its boundary. The length, width, and height directions of the cuboid bounding region can be parallel to the three coordinate axes of the preset 3D coordinate system. The cuboid bounding region can be voxelized to determine the voxel value of each voxel point. A voxel is the smallest unit of information in the three-dimensional space of a 3D object region. Each voxel has a voxel value, which represents the data information of the 3D object region at the location of that voxel.

[0086] The second sectional plane can be parallel to any preset coordinate axis in the preset 3D coordinate system that is not perpendicular to the first sectional plane; that is, the second sectional plane is not perpendicular to the first sectional plane. The second sectional plane can be parallel to or not parallel to the first sectional plane. Parallelism is preferable. On any second sectional plane, the intersection of the projection of the cuboid-enclosed region onto the second sectional plane (i.e., the sectional area) with the envelope line can form a sequence of intersection points. A closed polygonal region is constructed based on the sequence of intersection points. For example, the sequence of intersection points can be chained together to form the outline of the closed polygonal region, thus determining the closed polygonal region; or curve fitting can be performed based on the sequence of intersection points to obtain the outline of the closed polygonal region, thus determining the closed polygonal region. The voxel points contained within the closed polygonal region are assigned a first voxel value. The region formed by all voxel points with the first voxel value can represent the 3D object region where the target object is located. The voxel points in the cuboid-enclosed region that do not fall into the closed polygonal region on the current second sectional plane are assigned a second voxel value. The region formed by all voxel points with the second voxel value can represent the background region on the current second sectional plane.

[0087] When establishing the cuboid-enclosed region, an image resolution no lower than that of the original 3D image or the 2D cross-sectional image mapped onto the second cross-section can be preset. That is, the voxel array of the cuboid-enclosed region can be arranged in such a way that the voxels in the cuboid-enclosed region have a smaller or equal spacing value compared to the voxels in the original 3D image or the 2D cross-sectional image mapped onto the second cross-section. Furthermore, the voxel array can be stored in a 3D array using a binarized mask, that is, the voxel values ​​corresponding to each voxel in the voxel array are constructed as binary values ​​and stored in the 3D array. The row and column directions of the 3D array correspond one-to-one with the XYZ directions of the coordinate system used by the 3D image. The position of each voxel in the cuboid-enclosed region can be determined sequentially as follows: calculate the sequence of intersections between the second cross-section and the envelope of each layer, construct a closed polygon based on this sequence of intersections, and form a planar filled region. The voxels covered by the filled region within the current second cross-section can simultaneously be assigned a value to the 3D array, for example, a voxel value of 1. Batch processing of all layers allows for the complete recording of information about the 3D object region into a 3D array, thus completing the voxelization of the 3D object region. It can be understood that the constructed voxelized 3D object region can be considered a region mask of the 3D object region in the original 3D image.

[0088] In the above embodiments, the bounding region of the target object is determined by utilizing the value range of the target object in a preset coordinate system. The closed polygon region is then determined by using the intersection of each voxel point within the bounding region and the envelope. Subsequently, a value can be assigned to each pixel. When the target object has a high degree of shape irregularity, assigning values ​​to each voxel helps to accurately determine the 3D object region where the target object is located, greatly improving the accuracy of the method for determining the 3D marked region of the target object.

[0089] In some embodiments, determining the three-dimensional marked region containing the target object based on the contour set and the generated envelope further includes: calculating the gradient value of each voxel point within the cuboid-enclosed region; and determining voxel points with gradient values ​​greater than a preset gradient threshold as surface voxel points of the three-dimensional object region.

[0090] By comparing the gradient value at each pixel with a preset gradient threshold, voxel points located at the edges of the 3D object region can be identified, thus determining the surface voxel points of the 3D object region. The preset gradient threshold can be a threshold set theoretically or empirically to determine whether certain voxel points belong to the target object's edge. Voxel points with gradient values ​​greater than the threshold can be considered edges, while voxel points with gradient values ​​less than the threshold can be considered non-edges.

[0091] Morphological calculations typically process voxels across the entire image. To improve efficiency, based on the prior assumption that the 3D object region is a closed, simply connected region, the gradient value of the current voxel can be calculated. If this value exceeds a certain threshold (i.e., a preset gradient threshold), it can be identified as a surface voxel point within the 3D object region. This method controls the morphological dilation operation to be performed primarily on surface voxels, and only voxels whose values ​​change from 0 to 1 are marked, thus effectively distinguishing the safe region from the 3D object region.

[0092] In some embodiments, the method for determining the three-dimensional marked region of the target object further includes: projecting the three-dimensional marked region and the three-dimensional object region onto an observation surface (which may be referred to as the first observation surface) for display. The observation surface may be perpendicular to a preset target observation direction.

[0093] The observation plane can be any plane, which may or may not be parallel to any of the first and second tangent planes. After generating the 3D marked area, the user can observe the 3D marked area on the observation plane. The user can freely choose the direction of the observation plane according to actual needs; this invention does not impose any restrictions on this. The orientation matrix corresponding to the cuboid-enclosed area can be recorded. The orientation matrix can be established based on a preset coordinate system used for the cuboid-enclosed area. For example, the XYZ axes in the orientation matrix can be parallel to the XYZ axes of the preset coordinate system, that is, the direction cosine values ​​of each axis vector are defined to be the same as those in the preset coordinate system. At the same time, the orientation matrix can take the center point of the cuboid-enclosed area of ​​the target object as its origin. The 3D marked area and the 3D object area can be projected onto the observation plane for display based on the orientation matrix. On the observation plane, the first region contour obtained by projecting the 3D marked area and the second region contour obtained by projecting the 3D object area can be obtained.

[0094] For example, the following explanation uses a three-dimensional marked region as an example of a lesion region in the human body. Figure 5 A rendering of a three-dimensional marked area of ​​a lesion according to an embodiment of the present invention is shown. Figure 6a , 6b Figures 6c and 6d respectively illustrate schematic diagrams of a lesion on different sections according to an embodiment of the present invention. Figure 6a , 6bAs shown in Figures 6c and 6d, after obtaining the three-dimensional marked area of ​​the lesion, in response to user operation, the three-dimensional marked area can be projected onto any observation plane for display. The observation plane can be such as the coronal plane, sagittal plane, transverse plane, or any free section. Optionally, during projection, the three-dimensional object area can also be projected onto the observation plane for display. When projecting the three-dimensional marked area and the three-dimensional target object area onto the observation plane, the first region contour corresponding to the three-dimensional marked area and the second region contour corresponding to the three-dimensional object area can be specifically displayed. The first region contour represents the contour of the projected region of the three-dimensional marked area (i.e., the three-dimensional object area) on the observation plane before dilation processing. Figure 6a , 6b The inner solid lines in 6c and 6d. The second region contour is represented by the contour of the projected area of ​​the three-dimensional marked region on the observation plane after dilation, i.e. Figure 6a , 6b The outer solid lines in 6c and 6d.

[0095] In the above embodiments, users can observe the three-dimensional marked area and the three-dimensional object area of ​​the target object from different viewing planes, and can clearly distinguish the area before and after the dilation process, which meets the user's needs in different situations and effectively improves the flexibility and convenience of the user.

[0096] In some embodiments, the method for determining the three-dimensional marked region of a target object further includes: updating the contour set to obtain a new contour set, and re-executing the first determining step, the generating step, and the second determining step for the new contour set; wherein the updating operation includes one or more of the following operations: adding a new object contour to the contour set in response to an operation of adding an object contour to the contour set; deleting at least a portion of the object contours from the contour set in response to an operation of deleting an object contour to the contour set; generating one or more new object contours in response to a contour adjustment operation of one or more object contours in the contour set, and replacing the corresponding object contours in the contour set with the newly generated object contours to update the contour set.

[0097] For example, before updating the contour set to obtain a new contour set, the method for determining the three-dimensional marked region of the target object may further include: displaying a third region contour of the three-dimensional object region on any viewing plane (which may be called a second viewing plane) on the display interface, or displaying a third region contour and a fourth region contour of the three-dimensional marked region on the second viewing plane. The second viewing plane can be any plane, which may be parallel or non-parallel to the aforementioned first sectional plane, second sectional plane, and first viewing plane. The second viewing plane may be perpendicular to a preset target viewing direction. When the first viewing plane and the second viewing plane are not parallel, their corresponding target viewing directions are also not parallel. The second viewing plane may be, for example, a coronal plane, a sagittal plane, a cross-section, or any free sectional plane.

[0098] The display interface can be controlled to show the outline of the target object's third region on the second viewing plane, allowing the user to check the accuracy of the determined third region outline. If the determined third region outline is inaccurate, the user can delete or adjust the object outline on the second viewing plane. Of course, the user can also add new object outlines to the outline collection.

[0099] For example, when the target object has a high degree of irregularity in shape, under the same target viewing direction, the visualization software window can be lowered to view the object outline at each layer. If a significant deviation in the object outline is found when lowering the window to a certain cross-section, an update operation can be performed to appropriately adjust the object outline. The adjusted object outline is then added to the existing outline set. Based on the new outline set, the first determination step, the generation step, and the second determination step can be re-executed to redetermine the 3D marked area of ​​the target object.

[0100] Update operations can include adding a new first cross-section and the object contour formed within the first cross-section, or deleting object contours from the contour set that do not meet the requirements or have poor imaging effects. They can also adjust the object contours in the contour set to form new object contours. After adjusting the object contours, an update operation can be performed on the contour set.

[0101] In the above embodiments, adjustments and / or deletions and / or additions to object contours in the contour set are permitted, enabling the determination of the 3D object region containing the target object using a contour set that better meets user requirements. This allows the generated 3D object region to better suit user needs, effectively improving the user experience and the accuracy of the 3D object region results.

[0102] In some embodiments, determining multiple contour point groups based on a contour set (first determination step S120) includes: performing smooth resampling processing on multiple object contours in the contour set to obtain a contour point set uniquely corresponding to each object contour; determining multiple contour point groups based on the contour point sets corresponding to multiple object contours; wherein all contour points in each contour point set belong to different contour point groups in multiple contour point groups.

[0103] For example, smooth resampling can be implemented using one or more methods such as nearest neighbor interpolation, bilinear interpolation, and cubic convolution interpolation. Smoothing resampling can be performed separately for each object contour to obtain a denser distribution of contour points on each contour, thereby improving the accuracy of the subsequently determined envelope and consequently improving the accuracy of the determined 3D marked region.

[0104] After smoothing and resampling, contour points in any object contour and corresponding contour points in other object contours in the vertical direction can be combined to form a new contour point group.

[0105] In the above embodiments, by smoothing resampling, each object contour can have more and clearer contour points. The envelope formed by more contour points can more accurately and completely reflect the position of the target object, further improving the accuracy of the method for determining the three-dimensional marked region of the target object.

[0106] Figure 7 A schematic flowchart illustrating a method for determining a three-dimensional marked region of a target object according to an embodiment of the present invention is shown. Figure 7 This description uses a 3D CT / MR image as an example, with the target object being the lesion. Figure 7As shown, in step S701, the CT / MR image can first be displayed in sections. The displayed section can be any section, which serves as the first section. During section display, step S702 can be executed to move the parallel layer to the new section. By moving the parallel layer, different first sections of each layer can be displayed separately. For the currently displayed first section, step S703 can be executed to determine which contour generation method to use. If the automatic contour generation method is used, step S704 can be executed to automatically generate the curve (i.e., the object contour). If the manual contour generation method is used, step S705 can be executed to manually add contour sampling points (i.e., the labeled contour points described above). Subsequently, step S706 can be executed to fit the curve (i.e., the initial contour) based on the sampling points. Subsequently, step S707 can be optionally executed to edit and adjust the sampling points in response to the contour adjustment information input by the user, thereby adjusting the initial contour and obtaining the object contour. After steps S704 and S707, step S708 can be executed, which is to smooth and resample the curve (object contour). Then, step S709 is executed to determine whether the current curve is accepted. For example, if the user determines that the current object contour meets the requirements, a confirmation command (which can be called a second confirmation command) can be input. Upon receiving the second confirmation command input by the user, step S710 can be executed to store the current curve into the current curve set (i.e., the contour set described in this document). Then, step S711 can be executed to determine whether to construct an envelope. If not, the process can return to step S702, continue parallel layer shifting to obtain a new first cross-section, and continue executing step S703, etc., for the new first cross-section. If yes, the process can continue to step S712. In step S712, a vertical (i.e., along the direction perpendicular to the first cross-section) nearest neighbor search can be performed to obtain a contour point group. Then, step S713 is executed to perform curve fitting and connection based on the longitudinal contour point group. In step S714, the generated envelopes are obtained. Then, step S715 is executed to determine whether to construct a lesion region. If not, the process can return to step S702. If so, step S716 can be continued to perform mask calculation on the lesion area to determine the location of the lesion area (i.e., the three-dimensional object area described in this article). In step S717, a mask for the safe area can be further generated using the above-mentioned dilation scheme to obtain a complete mask for the dilated three-dimensional marked area. In step S718, the lesion cross-section is overlaid and displayed. In this step, the lesion contour and the contour of the entire three-dimensional marked area on any observation plane can be overlaid and displayed based on the masks of the lesion area and the three-dimensional marked area. The display effect can be referenced. Figures 6a-6d It should be noted that... Figure 7 The specific process of the method for determining the three-dimensional marked region of the target object shown is merely an example, and the present invention is not limited to this. Figure 7The process is illustrated. For example, as mentioned above, when manually determining the object outline, the user can also annotate the locations, and the object outline can be determined based on the pixel information of the annotated locations. Furthermore, some steps, such as S707 and S717, are optional and can be omitted.

[0107] According to another aspect of the present invention, a device for determining the three-dimensional marked region of a target object is also provided. Figure 8 A schematic block diagram of a three-dimensional marking region determination device for a target object according to an embodiment of the present invention is shown. Figure 8 As shown, the three-dimensional marking region determination device 800 for the target object includes an acquisition module 810, a first determination module 820, a generation module 830, and a second determination module 840.

[0108] The acquisition module 810 is used to acquire a contour set of the target object, the contour set including the object contours of the target object on multiple first cross-sections of a 3D image, the multiple first cross-sections being parallel to each other. The first determination module 820 is used to determine multiple contour point groups based on the contour set, wherein each contour point group includes a single contour point on each object contour in the contour set. For a first contour point on the current first cross-section and a second contour point on an adjacent first cross-section adjacent to the current first cross-section in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first cross-section, the target distance being a distance in a first target direction, the first target direction being a direction parallel to the multiple first cross-sections. The generation module 830 is used to generate an envelope based on each contour point group in the multiple contour point groups. The second determination module 840 is used to determine a 3D marked region containing the target object based on the contour set and the generated envelope.

[0109] For example, the generation module 830 includes: a fitting submodule for performing spline curve fitting based on each of a plurality of contour point groups to obtain an envelope corresponding to each contour point group.

[0110] For example, the acquisition module 810 includes: a first determining submodule, configured to determine an initial contour corresponding to any one of a plurality of first cross-sections; wherein the initial contour is an object contour; or, the acquisition module 810 further includes: a second determining submodule, configured to perform preset additional processing on the initial contour corresponding to the first cross-section to determine the object contour corresponding to the first cross-section; wherein, the first determining submodule includes: a detection unit, configured to perform automatic contour detection on the cross-section image corresponding to the first cross-section to obtain the initial contour corresponding to the first cross-section; or, the first determining submodule includes: a first acquisition unit. The first determination submodule includes: a second acquisition unit for acquiring multiple annotation contour points marked on the cross-sectional image corresponding to the first cross-section, and a first determination unit for determining the cross-sectional object region where the target object is located based on the pixel information at the position indicated by the object location information, and determining the contour of the cross-sectional object region as the initial contour corresponding to the first cross-section; or, the first determination submodule includes: a second acquisition unit for acquiring multiple annotation contour points marked on the cross-sectional image corresponding to the first cross-section, and a second determination unit for determining the initial contour corresponding to the first cross-section based on the multiple annotation contour points; wherein the cross-sectional image is determined based on the corresponding first cross-section and a three-dimensional image.

[0111] For example, the second determining unit includes: a sampling subunit, used to upsample or downsample multiple labeled contour points to obtain a new set of contour points; and a fitting subunit, used to perform spline curve fitting based on the new set of contour points to obtain the initial contour corresponding to the first cross-section.

[0112] For example, the second determining submodule includes: an adjustment unit, configured to adjust the initial contour corresponding to the first cross-section in response to input contour adjustment information, so as to obtain the object contour corresponding to the first cross-section.

[0113] For example, the second determining module 840 includes: a third determining submodule, used to determine the three-dimensional object region where the target object is located based on the contour set and the generated envelope; and an expansion submodule, used to expand outward with a preset thickness based on the surface voxel points of the three-dimensional object region, and obtain the expanded region as a three-dimensional marking region.

[0114] For example, the device 800 further includes a projection display module for projecting the three-dimensional marker area and the three-dimensional object area onto the viewing surface for display.

[0115] For example, the third determining submodule includes: a third determining unit, used to determine the value range of the target object on each coordinate axis in a preset three-dimensional coordinate system based on the contour set and the envelope; a building unit, used to build a cuboid-enclosed region of the target object based on the value range, wherein the length, width, and height of the cuboid-enclosed region are parallel to the three coordinate axes of the preset three-dimensional coordinate system in a one-to-one correspondence; a fourth determining unit, used to determine the sequence of intersection points between the cuboid-enclosed region and the envelope on each of a plurality of second sectional planes parallel to the preset coordinate axes in the preset three-dimensional coordinate system; and a construction unit. The system is used to construct a closed polygon region based on the intersection sequence; the assignment unit is used to assign the voxel points contained within the closed polygon region to the first voxel value, and assign the voxel points outside the closed polygon region in the sectional region to the second voxel value; the fifth determination unit is used to determine the region where the voxel points within the cuboid-enclosed region have voxel values ​​equal to the first voxel value as the three-dimensional object region; wherein, the preset coordinate axis is not perpendicular to the multiple first sectional planes, the spacing between the multiple second sectional planes in the second target direction is equal to the spacing between the voxel points in the cuboid-enclosed region in the second target direction, and the second target direction is a direction perpendicular to the multiple second sectional planes.

[0116] For example, the second determining module 840 further includes: a calculation submodule for calculating the gradient value of each voxel point within the cuboid-enclosed region; and a fourth determining submodule for determining voxel points whose gradient values ​​are greater than a preset gradient threshold as surface voxel points of the three-dimensional object region.

[0117] For example, the apparatus 800 further includes: an update module, configured to perform an update operation on the contour set to obtain a new contour set, and restart the first determining module 820, the generating module 830, and the second determining module 840 for the new contour set; wherein the update operation includes one or more of the following operations: adding a new object contour to the contour set in response to an operation of adding an object contour to the contour set; deleting at least a portion of the object contours from the contour set in response to an operation of deleting an object contour to the contour set; generating one or more new object contours in response to a contour adjustment operation on one or more object contours in the contour set, and replacing the corresponding object contours in the contour set with the newly generated object contours to update the contour set.

[0118] For example, the first determining module 820 includes: a resampling submodule, used to perform smooth resampling processing on multiple object contours in the contour set to obtain a set of contour points uniquely corresponding to each object contour; and a fifth determining submodule, used to determine multiple contour point groups based on the set of contour points corresponding to the multiple object contours; wherein all contour points in each contour point set belong to different contour point groups in the multiple contour point groups.

[0119] According to another aspect of the present invention, an electronic device is also provided. Figure 9 A schematic block diagram of an electronic device according to an embodiment of the present invention is shown. Figure 9 As shown, the electronic device 900 includes a processor 910 and a memory 920. The memory 920 stores computer program instructions, which are executed by the processor 910 to perform the above-described method for determining the three-dimensional marked region of the target object.

[0120] Furthermore, according to another aspect of the present invention, a storage medium is also provided. Program instructions are stored on the storage medium. When the program instructions are executed by a computer or processor, the computer or processor performs the corresponding steps of the method for determining the three-dimensional marked region of the target object described in the embodiments of the present invention, and is used to implement the corresponding module of the method for determining the three-dimensional marked region of the target object described in the embodiments of the present invention or the corresponding module in the electronic device described above. The storage medium may, for example, include a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disc read-only memory (CD-ROM), a USB memory, or any combination of the above storage media. A computer-readable storage medium may be any combination of one or more computer-readable storage media.

[0121] According to another aspect of the present invention, a computer program product is also provided, including computer program instructions, which, when executed, are used to perform the above-described method for determining the three-dimensional marked region of the target object.

[0122] Those skilled in the art can understand the specific implementation scheme of the above electronic device and storage medium by reading the relevant description of the method for determining the three-dimensional marked region of the target object. For the sake of brevity, it will not be described in detail here.

[0123] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of the invention. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of the invention. All such changes and modifications are intended to be included within the scope of the invention as claimed in the appended claims.

[0124] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0125] In the several embodiments provided by this invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed.

[0126] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0127] Similarly, it should be understood that, in order to streamline the invention and aid in understanding one or more of the various aspects of the invention, features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof in the description of exemplary embodiments of the invention. However, this approach should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the corresponding claims, its inventive point lies in solving the corresponding technical problem with fewer features than all of those in a single disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.

[0128] Those skilled in the art will understand that, apart from the mutual exclusion of features, all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or elements of any method or apparatus so disclosed may be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0129] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments can be used in any combination.

[0130] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some modules in the three-dimensional marking region determination device for a target object according to embodiments of the present invention. The present invention can also be implemented as a device program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0131] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0132] The above description is merely a specific embodiment of the present invention or an explanation of that embodiment. The scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for determining a three-dimensional marked region of a target object, characterized in that, The method includes: Acquisition Steps: Acquire the contour set of the target object, the contour set including the object contours of the target object on multiple first cross-sections of the three-dimensional image, the multiple first cross-sections being cross-sections that are parallel to each other; First determination step: Based on the contour set, determine multiple contour point groups, wherein each contour point group includes a single contour point on each object contour in the contour set. For a first contour point on the current first sectional plane and a second contour point on an adjacent first sectional plane adjacent to the current first sectional plane in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first sectional plane. The target distance is the distance in a first target direction, and the first target direction is a direction parallel to the multiple first sectional planes. Generation step: Generate an envelope based on each of the plurality of contour point groups; The second determination step: Based on the contour set and the generated envelope, determine the three-dimensional marked region containing the target object; The step of determining the 3D marked region containing the target object based on the contour set and the generated envelope includes: The three-dimensional object region where the target object is located is determined based on the contour set and the generated envelope. The three-dimensional marked region is obtained based on the three-dimensional object region; The step of determining the 3D object region where the target object is located based on the contour set and the generated envelope includes: The range of values ​​for the target object on each coordinate axis in the preset three-dimensional coordinate system is determined based on the contour set and the envelope. Based on the value range, a cuboid-enclosed region of the target object is established, and the length, width, and height of the cuboid-enclosed region are parallel to the three coordinate axes of the preset three-dimensional coordinate system in a one-to-one correspondence. For each of the multiple second tangents parallel to the preset coordinate axes in the preset three-dimensional coordinate system Determine the sequence of intersection points between the cross-sectional region of the cuboid-enclosed region on the second cross-section and the envelope line; Construct a closed polygonal region based on the intersection point sequence; The voxel points contained within the closed polygon region are assigned a first voxel value, and the voxel points located outside the closed polygon region in the sectional region are assigned a second voxel value. The region containing the voxel point whose voxel value is equal to the first voxel value within the area enclosed by the cuboid is defined as the three-dimensional object region. Wherein, the preset coordinate axis is not perpendicular to the plurality of first cut surfaces, the spacing between the plurality of second cut surfaces in the second target direction is equal to the voxel spacing of the cuboid-enclosed region in the second target direction, and the second target direction is a direction perpendicular to the plurality of second cut surfaces.

2. The method as described in claim 1, characterized in that, The step of generating an envelope based on each of the plurality of contour point groups includes: Spline curve fitting is performed on each of the plurality of contour point groups to obtain the envelope corresponding to each contour point group.

3. The method as described in claim 1, characterized in that, The process of obtaining the contour set of the target object includes: For any one of the plurality of first cut surfaces, determine the initial contour corresponding to that first cut surface; Wherein, the initial contour is the object contour, or, the acquisition of the contour set of the target object further includes: for any one of the plurality of first facets, performing a preset additional processing on the initial contour corresponding to the first facet to determine the object contour corresponding to the first facet; The determination of the initial contour corresponding to the first cross-section includes: Automatic contour detection is performed on the cross-sectional image corresponding to the first cross-section to obtain the initial contour corresponding to the first cross-section; or, Obtain the object location information marked on the cross-sectional image corresponding to the first cross-section; determine the cross-sectional object region where the target object is located based on the pixel information at the position indicated by the object location information; and determine the contour of the cross-sectional object region as the initial contour corresponding to the first cross-section; or... Obtain multiple annotation contour points marked on the cross-sectional image corresponding to the first cross-section, and determine the initial contour corresponding to the first cross-section based on the multiple annotation contour points; The cross-sectional image is determined based on the corresponding first cross-section and the three-dimensional image.

4. The method as described in claim 3, characterized in that, Determining the initial contour corresponding to the first cross-section based on the plurality of marked contour points includes: Upsampling or downsampling of the multiple labeled contour points yields a new set of contour points; Based on the new set of contour points, spline curve fitting is performed to obtain the initial contour corresponding to the first cross-section.

5. The method as described in claim 3, characterized in that, The preset additional processing includes: In response to the input contour adjustment information, the initial contour corresponding to the first cross-section is adjusted to obtain the object contour corresponding to the first cross-section.

6. The method according to any one of claims 1-5, characterized in that, Obtaining the three-dimensional marked region based on the three-dimensional object region includes: The surface voxel points of the three-dimensional object region are used as a reference to expand outward by a preset thickness, and the expanded region is used as the three-dimensional marking region.

7. The method as described in claim 6, characterized in that, The method further includes: The three-dimensional marked area and the three-dimensional object area are projected onto the observation surface for display.

8. The method according to any one of claims 1-5, characterized in that, The step of determining the 3D marked region containing the target object based on the contour set and the generated envelope further includes: Calculate the gradient value of each voxel point within the region enclosed by the cuboid; Voxel points whose gradient values ​​are greater than a preset gradient threshold are identified as surface voxel points of the three-dimensional object region.

9. The method according to any one of claims 1-5, characterized in that, The method further includes: The contour set is updated to obtain a new contour set, and the first determining step, the generating step, and the second determining step are re-executed for the new contour set. The update operation includes one or more of the following operations: In response to an operation of adding an object contour to the contour set, a new object contour is added to the contour set. In response to a deletion operation on object contours of the contour set, at least a portion of the object contours are deleted from the contour set. In response to a contour adjustment operation on one or more object contours in the contour set, one or more new object contours are generated, and the corresponding object contours in the contour set are replaced with the newly generated object contours to update the contour set.

10. The method according to any one of claims 1-5, characterized in that, The step of determining multiple contour point groups based on the contour set includes: Smoothing resampling is performed on multiple object contours in the contour set to obtain a set of contour points that uniquely corresponds to each object contour. The plurality of contour point groups are determined based on the set of contour points corresponding to the contours of the plurality of objects; In each contour point set, all contour points belong to different contour point groups within the multiple contour point groups.

11. A device for determining a three-dimensional marked region of a target object, characterized in that, The device includes: The acquisition module is used to acquire a set of contours of a target object, the set of contours including the object contours of the target object on multiple first cross-sections of a three-dimensional image, the multiple first cross-sections being cross-sections that are parallel to each other; The first determining module is used to determine multiple contour point groups based on the contour set, wherein each contour point group includes a single contour point on each object contour in the contour set, and for a first contour point on the current first sectional plane and a second contour point on an adjacent first sectional plane adjacent to the current first sectional plane in each contour point group, the first contour point is the contour point with the closest target distance to the second contour point among the contour points on the current first sectional plane, and the target distance is the distance in a first target direction, which is a direction parallel to the multiple first sectional planes; The generation module is used to generate an envelope based on each of the plurality of contour point groups; The second determining module is used to determine a three-dimensional marked region containing the target object based on the contour set and the generated envelope. The second determining module includes: The third determining submodule is used to determine the three-dimensional object region where the target object is located based on the contour set and the generated envelope; the obtaining submodule is used to obtain the three-dimensional marking region based on the three-dimensional object region; The third determining submodule includes: The third determining unit is used to determine the value range of the target object on each coordinate axis in the preset three-dimensional coordinate system based on the contour set and the envelope. A unit is established to establish a cuboid-enclosed region of the target object based on the value range, wherein the length, width, and height of the cuboid-enclosed region are parallel to the three coordinate axes of the preset three-dimensional coordinate system in a one-to-one correspondence. The fourth determining unit is used to determine, for each of the multiple second tangents parallel to the preset coordinate axes in the preset three-dimensional coordinate system, the sequence of intersection points between the tangent region of the cuboid-enclosed region on the second tangent and the envelope line. Construction unit, used to construct a closed polygon region based on the intersection sequence; The assignment unit is used to assign a first voxel value to voxel points contained within the closed polygon region, and to assign a second voxel value to voxel points located outside the closed polygon region in the sectional region. The fifth determining unit is used to determine the region where the voxel point with a voxel value equal to the first voxel value is located within the region enclosed by the cuboid as the three-dimensional object region; Wherein, the preset coordinate axis is not perpendicular to the plurality of first cut surfaces, the spacing between the plurality of second cut surfaces in the second target direction is equal to the voxel spacing of the cuboid-enclosed region in the second target direction, and the second target direction is a direction perpendicular to the plurality of second cut surfaces.

12. An electronic device comprising a processor and a memory, characterized in that, The memory stores computer program instructions, which, when executed by the processor, are used to perform the method for determining the three-dimensional marked region of a target object as described in any one of claims 1-9.

13. A storage medium storing a computer program / instruction, characterized in that, The computer program / instructions are used to execute the method for determining the three-dimensional marked region of the target object as described in any one of claims 1-9.

14. A computer program product comprising computer program instructions, characterized in that, The computer program instructions, when executed, are used to perform the method for determining the three-dimensional marked region of the target object as described in any one of claims 1-9.

Images