CN117952990A - Three-dimensional target object segmentation method based on active contour model - Google Patents
Three-dimensional target object segmentation method based on active contour model Download PDFInfo
- Publication number
- CN117952990A CN117952990A CN202410120048.XA CN202410120048A CN117952990A CN 117952990 A CN117952990 A CN 117952990A CN 202410120048 A CN202410120048 A CN 202410120048A CN 117952990 A CN117952990 A CN 117952990A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- contour
- plane
- point
- segmentation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000011218 segmentation Effects 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims abstract description 39
- 238000009792 diffusion process Methods 0.000 claims abstract description 8
- 238000013507 mapping Methods 0.000 claims abstract description 4
- 238000001914 filtration Methods 0.000 claims description 6
- 238000001514 detection method Methods 0.000 claims description 3
- 238000012804 iterative process Methods 0.000 claims description 3
- 238000004458 analytical method Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 20
- 238000002591 computed tomography Methods 0.000 description 4
- 238000002059 diagnostic imaging Methods 0.000 description 4
- 238000003709 image segmentation Methods 0.000 description 4
- 238000012014 optical coherence tomography Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000002595 magnetic resonance imaging Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- 210000003484 anatomy Anatomy 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
- 238000012800 visualization Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a three-dimensional target object segmentation method based on an active contour model, which comprises the following steps: step 1: acquiring original three-dimensional medical image data; step 2: selecting a point inside the three-dimensional target area as a target seed point; step 3: searching an initial three-dimensional contour point set based on a spherical diffusion model; step 4: mapping the initial three-dimensional contour point set to two perpendicular plane sets; step 5: performing region segmentation on the two-dimensional plane by adopting an active contour model; step 6: and integrating all the two-dimensional region segmentation results to obtain a three-dimensional segmentation result. The invention can conveniently and rapidly segment the three-dimensional target object, and ensure the accuracy of the three-dimensional target object, so as to help a user to segment and extract the three-dimensional result better and perform subsequent analysis. The method is wide in applicability, is not limited to three-dimensional medical images, and is also suitable for other types of three-dimensional images.
Description
Technical Field
The invention belongs to the technical field of medical image processing, and relates to a three-dimensional target object segmentation method based on an active contour model.
Background
Medical imaging technology has evolved tremendously over the last three decades, allowing one to acquire images of human anatomy and tissue function in space and time. The advent of medical imaging technology has enabled researchers and doctors to obtain patient-underlying disease information without contact. However, since medical images have special imaging procedures and image characteristics, a specific image processing method is required.
Medical image data is typically three-dimensional, such as CT (computed tomography) and MRI (magnetic resonance imaging) and OCT (optical coherence tomography). Medical imaging has long been an important branch of the medical field, providing doctors with information about the tissues and organs of patients, and three-dimensional medical imaging segmentation refers to the automatic or semi-automatic segmentation and extraction of a target region of interest in a medical image. The medical image segmentation has important application value in the medical field, and can be used for disease diagnosis, operation planning, treatment monitoring and other aspects.
The medical image has the characteristics of complexity, diversity and the like, has certain noise, and the edges of organs in the image are also locally unclear, so that the three-dimensional target object in the medical image is difficult to extract. Several methods commonly used at present are a deep learning method, which can realize automatic segmentation of medical images, but a large amount of labeling data is needed for training; the region growing method is characterized in that the region is gradually expanded by combining pixels similar to the seed points based on the seed points, and the method has a good image effect with clear edges and consistent gray scales in the region; the segmentation method based on the threshold value is used for segmenting the image into a target and a background according to the threshold value of the pixel gray value. This approach is simple and fast, but may not accurately segment the target in processing complex images.
The existing image segmentation methods have the defects in different aspects.
Disclosure of Invention
In order to solve the problems, the invention discloses a three-dimensional target object segmentation method based on an active contour model.
The specific scheme is as follows:
a three-dimensional target object segmentation method based on an active contour model comprises the following steps:
step 1: acquiring original three-dimensional medical image data;
step 2: selecting a point inside the three-dimensional target area as a target seed point;
step 3: searching an initial three-dimensional contour point set based on a spherical diffusion model;
step 4: mapping the initial three-dimensional contour point set to two perpendicular plane sets;
step 5: performing region segmentation on the two-dimensional plane by adopting an active contour model;
step 6: and integrating all the two-dimensional region segmentation results to obtain a three-dimensional segmentation result.
Further, the three-dimensional medical image data in the step 1 is any conventional three-dimensional medical image, such as CT/MR/OCT, etc.
Further, in the step 2, a point inside the three-dimensional target area is selected as the target seed point, and the point should be a point near the center inside the three-dimensional target area.
Further, in the step 3, an initial three-dimensional contour point set is found based on a spherical diffusion model, and the specific method comprises the following steps:
Step 3.1: 2, taking the target seed point in the step as a center, and transmitting a ray set outwards in a spherical mode;
step 3.2: acquiring an image voxel set corresponding to each ray to obtain one-dimensional data;
step 3.3: filtering the one-dimensional data in the step 3.2;
Step 3.4: carrying out gradient detection on the one-dimensional data filtered in the step 3.3 to obtain one-dimensional data boundary points, and recording gradient values;
step 3.5: and obtaining all spherical ray boundary points to form an initial contour point set of the three-dimensional target area.
Further, in the step 4, the initial three-dimensional contour is mapped to two perpendicular plane sets, and the specific method is as follows:
Step 4.1: the coordinates of the three-dimensional contour point set are rounded downwards;
step 4.2: obtaining a bounding box from the rounded point set,
Represented by (min x,maxx,miny,maxy,minz,maxz);
Step 4.3: expanding along the X-axis direction to obtain a series of plane point sets, namely, the X values of all points on a plane are the same;
step 4.4: expanding along the Y-axis direction to obtain a series of plane point sets, namely, the Y values of all points on a plane are the same;
Step 4.5: the point sets in the two plane sets are distributed according to gradient, and noise data are filtered;
Step 4.6: performing curve fitting on the filtered plane point set, wherein the obtained curve is the current plane region segmentation initial contour;
Step 4.7: and (4) obtaining all initial contours of the three-dimensional target area for all planes in the mode of step 4.6.
Further, in the step 5, an active contour model is adopted to segment the two-dimensional plane, and an active contour algorithm is based on the idea of curve evolution to realize segmentation by automatically searching the boundary of the region of interest in the image; the specific method comprises the following steps:
Step 5.1: initializing a contour, and acquiring the initial contour of a certain plane acquired in the step 4;
Step 5.2: defining an energy function, the energy function generally comprising an internal energy and an external energy, with a curvature energy term as the internal energy function; the external energy function is an energy term based on the image gradient;
Step 5.3: optimizing an energy function in an iterative mode, and adjusting the shape of the profile; in the iterative process, updating the position of the contour according to the minimum value of the energy function;
Step 5.4: stop conditions are defined, conditions for stopping the iteration are defined, such as the number of iterations reaching a threshold, the profile converging or the change in the energy function being less than a certain threshold, etc.
Further, in the step 6, all the two-dimensional region segmentation results are integrated to obtain a three-dimensional segmentation result, specifically, the results obtained by dividing all the planes in the step 5 are combined, and the obtained results are the segmentation results of the three-dimensional target region.
The invention has the beneficial effects that: by applying the technical scheme of the invention, the three-dimensional target object can be conveniently and rapidly segmented, and the accuracy of the three-dimensional target object is ensured, so that a user can be helped to better segment and extract the three-dimensional result, and the follow-up analysis can be carried out. The method is wide in applicability, is not limited to three-dimensional medical images, and is also suitable for other types of three-dimensional images.
Drawings
FIG. 1 is a schematic flow chart of the method of the present invention.
FIG. 2 is a plot of seed points selected to be near the center of the target area in accordance with the present invention.
FIG. 3 is a graph of a spherical diffusion model in accordance with the present invention.
FIG. 4 is a schematic diagram of the calculation of the point set on the sphere in the present invention.
FIG. 5 is a graph showing the effect of curve fitting based on a point set in the present invention.
FIG. 6 is a schematic view of the initial contour of the target area according to the present invention.
FIG. 7 is a schematic view of a target region contour generated based on an active contour model in accordance with the present invention.
Fig. 8 is a schematic diagram of a three-dimensional object visualization result obtained based on active contour model segmentation in the present invention.
Detailed Description
The present invention is further illustrated in the following drawings and detailed description, which are to be understood as being merely illustrative of the invention and not limiting the scope of the invention.
As shown in fig. 1, the present invention provides a three-dimensional object segmentation method based on an active contour model, which includes the following steps:
Step 1: original three-dimensional image data is acquired. Can be any conventional three-dimensional medical image such as CT/MR/OCT. In addition, the method is also applicable to three-dimensional medical image data as well as other types of three-dimensional image data. This implementation example selects an MRI image as an example.
Step 2: and selecting a point in the three-dimensional target area as a target seed point, wherein the point is selected, a three-dimensional target area object can be determined through three-dimensional image preview software, and the three-dimensional target area object is positioned to the point in the three-dimensional target area, and the point is a point in the three-dimensional target area, which is close to the center. As shown in fig. 2.
Step 3: searching an initial three-dimensional contour point set based on a spherical diffusion model, wherein the spherical diffusion model is shown in fig. 3, and the specific method comprises the following steps:
Step 3.1: and (3) taking the seed point in the step (2) as a center, and emitting the ray set outwards in a spherical mode. Step 3.2: the sphere can be represented by a parametric equation, as shown in fig. 4, assuming that the sphere center is (x 0, y0, z 0), the radius is r, and the parameters θ and Φ are respectively represented as a polar angle and an azimuth angle, then any point (x, y, z) on the sphere can be represented by the following parametric equation:
z＝z0+r*cosθ
Step 3.3: and acquiring an image voxel set corresponding to each ray to obtain one-dimensional data, namely extracting one-dimensional image information along the ray direction, wherein the extraction mode can be nearest interpolation or other interpolation modes.
Step 3.4: filtering the one-dimensional data in the step 3.3 to eliminate noise points in the image, wherein a one-dimensional mean value filtering mode can be adopted, and a filtering core can be 1*3;
Step 3.5: and (3) carrying out gradient detection on the one-dimensional data filtered in the step (3.4) to obtain one-dimensional data boundary points, and recording gradient values. The one-dimensional gradient can be solved by a differential mode, such as a center difference method, a forward difference method and the like. After solving the gradient, the gradient value can be compared and analyzed, the first peak point is taken as the boundary point of the one-dimensional data, and the current gradient value is recorded.
Step 3.6: repeating the operation in the step 3.5 for all spherical rays, obtaining all boundary points, and forming an initial contour point set of the three-dimensional target area.
Step 4: the initial three-dimensional contour maps to two perpendicular sets of planes, transforming the three-dimensional segmentation into a two-dimensional segmentation. The three-dimensional point sets may be projected in either a horizontal or vertical direction to form two perpendicular plane sets, with the point set on each plane representing the initial contour point of the target region on that plane. By adopting two perpendicular plane sets, the effect that some edge areas are poorly segmented only in one direction can be compensated. The specific method comprises the following steps:
Step 4.1: the three-dimensional contour point set coordinates are rounded downwards, and before rounding, the point coordinates are non-shaped and need to be converted into pixel coordinates, so that the point set coordinates are rounded in advance.
Step 4.2: obtaining a bounding box from the rounded point set,
Indicated by (min x,maxx,miny,maxy,minz,maxz).
Step 4.3: and expanding along the X-axis direction to obtain a series of plane point sets, namely, all points on the plane have the same X value, traversing all point coordinates in the point sets in actual operation, classifying all points with the same X value into one type, namely, representing the initial contour point set of the plane, and all points are in the bounding box.
Step 4.4: and (3) expanding along the Y-axis direction to obtain a series of plane point sets, wherein Y values of all points on a plane are the same, and the practical implementation mode can be as described in the step 4.3.
Step 4.5: and filtering noise data according to gradient distribution for point sets in the two plane sets. The point sets in each plane represent the initial contour point set of the plane, and all the point sets in the current plane can be subjected to statistical gradient distribution, if the histogram distribution is adopted to express the gradient distribution, the data at two ends of the histogram distribution can be cut off, and the data in the middle area of the distribution can be reserved as the filtered contour point set. For example, 20% of data can be truncated at the front and rear ends, respectively, according to the distribution, and 60% of data in the middle area can be reserved.
Step 4.6: and performing curve fitting on the filtered plane point set to obtain a curve which is the initial contour of the current plane region segmentation, wherein the curve fitting can be performed in a B-spline curve fitting or Bezier curve and other modes. The result of fitting the curve based on the point set is shown in fig. 5. The initial contour is superimposed on the actual image and displayed as a fusion diagram, as shown in fig. 6.
Step 5: and dividing the target area by adopting an active contour algorithm. The target area here is a two-dimensional target area. In addition, the active contour algorithm is an image segmentation method based on energy minimization. The basic idea is to place an initial contour in the image and to gradually approach the target boundary by adjusting the shape and position of the contour, thereby achieving image segmentation. The active contour algorithm is an algorithm idea and has a plurality of specific algorithm forms, and takes a Chan-Vese (CV) level set segmentation algorithm as an example, and the level set algorithm has the advantage that the level set algorithm can adaptively adjust the shape and the topological structure of the contour, so that the level set algorithm has higher precision and robustness when processing images with complex shapes and irregular boundaries. The specific implementation steps are as follows:
Step 5.1: initializing a contour, namely taking a curve fitted by two-dimensional plane contour points acquired from the three-dimensional initial contour points in the step 4 as an initial contour, wherein the initial contour can relatively describe the contour of the target area at the moment. Step 5.2: defining an energy function, the energy function generally comprising an internal energy and an external energy, with a curvature energy term as the internal energy function; the external energy function is an energy term based on image gradients
The energy function is as follows:
Where E denotes the energy function, c1 and c2 are two different gray values, Is a level set function, μ, ν, λ1, and λ2 are weight parameters, I is an input image, and R1 and R2 are divided regions.
The length of the representation profile can be obtained by calculating the gradient modulus of the level set function.
Area (R 1,R2) represents the Area of the divided region, and can be obtained by calculating the number of pixels whose level set function is greater than 0 and less than 0.
RegionTerm (I, R) represent energy terms based on the brightness of the image, which can be obtained by calculating the difference square of the gray-scale average value and the gray-scale value of the image in the divided area. This term is used to push the contour to move towards areas of the image where the grey level changes a lot.
It should be noted that the image here is a two-dimensional image slice of the position corresponding to the contour point set of the current two-dimensional plane. For example, assuming that the current point set is a point set at x=k, the image slice should be a two-dimensional slice of three-dimensional image data corresponding to x=k.
Step 5.3: and optimizing the energy function in an iterative mode, and adjusting the shape of the profile. In the iterative process, the position of the contour is updated according to the minimum value of the energy function.
Step 5.4: stop conditions are defined, conditions for stopping the iteration are defined, such as the number of iterations reaching a threshold, the change in the energy function being less than a certain threshold, etc. One possible parameter setting scheme is: setting the maximum iteration number as 200; setting the root mean square error threshold value of iteration stop to be 0.03; setting the coefficient of the curvature item to be 1; setting the coefficient of the propagation term to be 1; the coefficient of the advection term is set to 1.
Step 5.5: and extracting a pixel region in the contour as a target region segmentation result of the current plane, wherein the contour of the target region obtained by segmentation based on the active contour model is shown in fig. 7 and is close to the real boundary of the target region.
Step 5.6: in order to obtain finer segmentation results, it is conceivable to perform a suitable region growth in the edge region on the basis of the results in step 5.5, which is advantageous for segmentation of certain very irregular target regions.
Step 6: integrating all the two-dimensional region segmentation results to obtain a three-dimensional segmentation result, wherein the specific method comprises the following steps:
step 6.1: and (5) taking a union set to form a three-dimensional segmentation model based on all the two-dimensional plane segmentation results obtained by segmentation in the step (5).
Step 6.2: the three-dimensional model is smoothed by adopting an average method, a weighted average method, a curved surface fitting method and the like.
Step 6.3: the result of the three-dimensional object segmentation is obtained, and the result of the three-dimensional object segmentation is visualized as shown in fig. 8.
Claims (7)
1. The three-dimensional target object segmentation method based on the active contour model is characterized by comprising the following steps of:
step 1: acquiring original three-dimensional medical image data;
step 2: selecting a point inside the three-dimensional target area as a target seed point;
step 3: searching an initial three-dimensional contour point set based on a spherical diffusion model;
step 4: mapping the initial three-dimensional contour point set to two perpendicular plane sets;
step 5: performing region segmentation on the two-dimensional plane by adopting an active contour model;
step 6: and integrating all the two-dimensional region segmentation results to obtain a three-dimensional segmentation result.
2. The method for segmenting a three-dimensional object based on an active contour model according to claim 1, wherein the three-dimensional medical image data in step 1 is any conventional three-dimensional medical image.
3. The method according to claim 2, wherein in the step 2, a point inside the three-dimensional target area is selected as the target seed point, and the point is a point near the center inside the three-dimensional target area.
4. The method for segmenting the three-dimensional target object based on the active contour model according to claim 3, wherein the initial three-dimensional contour point set is found based on the spherical diffusion model in the step 3, and the specific method comprises the following steps:
Step 3.1: 2, taking the target seed point in the step as a center, and transmitting a ray set outwards in a spherical mode;
step 3.2: acquiring an image voxel set corresponding to each ray to obtain one-dimensional data;
step 3.3: filtering the one-dimensional data in the step 3.2;
Step 3.4: carrying out gradient detection on the one-dimensional data filtered in the step 3.3 to obtain one-dimensional data boundary points, and recording gradient values;
step 3.5: and obtaining all spherical ray boundary points to form an initial contour point set of the three-dimensional target area.
5. The method for three-dimensional object segmentation based on the active contour model according to claim 4, wherein the mapping of the initial three-dimensional contour to two perpendicular plane sets in step 4 is as follows:
Step 4.1: the coordinates of the three-dimensional contour point set are rounded downwards;
step 4.2: obtaining a bounding box from the rounded point set,
Represented by (min x,maxx,miny,maxy,minz,maxz);
Step 4.3: expanding along the X-axis direction to obtain a series of plane point sets, namely, the X values of all points on a plane are the same;
step 4.4: expanding along the Y-axis direction to obtain a series of plane point sets, namely, the Y values of all points on a plane are the same;
Step 4.5: the point sets in the two plane sets are distributed according to gradient, and noise data are filtered;
Step 4.6: performing curve fitting on the filtered plane point set, wherein the obtained curve is the current plane region segmentation initial contour;
Step 4.7: and (4) obtaining all initial contours of the three-dimensional target area for all planes in the mode of step 4.6.
6. The method for segmenting the three-dimensional target object based on the active contour model according to claim 5, wherein in the step 5, the two-dimensional plane is segmented by adopting the active contour model, and the active contour algorithm is based on the idea of curve evolution, and the segmentation is realized by automatically searching the boundary of the region of interest in the image; the specific method comprises the following steps:
Step 5.1: initializing a contour, and acquiring the initial contour of a certain plane acquired in the step 4;
Step 5.2: defining an energy function, wherein the energy function comprises internal energy and external energy, and curvature energy term is used as the internal energy function; the external energy function is an energy term based on the image gradient;
Step 5.3: optimizing an energy function in an iterative mode, and adjusting the shape of the profile; in the iterative process, updating the position of the contour according to the minimum value of the energy function;
step 5.4: stop conditions are defined, and iteration stop conditions are defined.
7. The method for segmenting the three-dimensional target object based on the active contour model according to claim 6, wherein in the step 6, all the two-dimensional region segmentation results are integrated to obtain a three-dimensional segmentation result, specifically, the results obtained by dividing all the planes in the step 5 are combined, and the obtained result is the segmentation result of the three-dimensional target region.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410120048.XA CN117952990A (en) | 2024-01-29 | 2024-01-29 | Three-dimensional target object segmentation method based on active contour model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410120048.XA CN117952990A (en) | 2024-01-29 | 2024-01-29 | Three-dimensional target object segmentation method based on active contour model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117952990A true CN117952990A (en) | 2024-04-30 |
Family
ID=90791839
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410120048.XA Pending CN117952990A (en) | 2024-01-29 | 2024-01-29 | Three-dimensional target object segmentation method based on active contour model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117952990A (en) |
-
2024
- 2024-01-29 CN CN202410120048.XA patent/CN117952990A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11776216B2 (en) | System and method for extracting a region of interest from volume data | |
| Sun et al. | Automated 3-D segmentation of lungs with lung cancer in CT data using a novel robust active shape model approach | |
| US8750585B2 (en) | Method and system for segmenting a liver object in an image | |
| CN108038862B (en) | Interactive medical image intelligent segmentation modeling method | |
| WO2019000455A1 (en) | Method and system for segmenting image | |
| JP6539303B2 (en) | Transforming 3D objects to segment objects in 3D medical images | |
| US20030068074A1 (en) | Computer system and a method for segmentation of a digital image | |
| CA2986590C (en) | Surface modeling of a segmented echogenic structure for detection and measurement of anatomical anomalies | |
| CN110992377B (en) | Image segmentation method, device, computer-readable storage medium and equipment | |
| CN107680110B (en) | Inner ear three-dimensional level set segmentation method based on statistical shape model | |
| CN110675464A (en) | Medical image processing method and device, server and storage medium | |
| CN116580068B (en) | Multi-mode medical registration method based on point cloud registration | |
| CN112136157A (en) | Method, system and computer program for segmenting tooth pulp regions from images | |
| JP2021529622A (en) | Method and computer program for segmentation of light interference tomography images of the retina | |
| Jaffar et al. | Anisotropic diffusion based brain MRI segmentation and 3D reconstruction | |
| CN110570430B (en) | Orbital bone tissue segmentation method based on volume registration | |
| CN117952990A (en) | Three-dimensional target object segmentation method based on active contour model | |
| Zhuo et al. | Local adaptive segmentation algorithm for 3-D medical image based on robust feature statistics | |
| CN113706687A (en) | Nose environment modeling method and device for path planning | |
| Reska et al. | Fast 3D segmentation of hepatic images combining region and boundary criteria | |
| CN111489434A (en) | Medical image three-dimensional reconstruction method based on three-dimensional graph cut | |
| CN117495893B (en) | Skull peeling method based on active contour model | |
| US20240112331A1 (en) | Medical Image Data Processing Technique | |
| Puyuelo Citoler | The face as a window to the brain: Development of a pipeline for facial segmentation from magnetic resonance scans | |
| CN108492304B (en) | Medical image segmentation method based on multi-directional contour |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |