CN111402277A - Object outer contour segmentation method and device for medical image - Google Patents
Object outer contour segmentation method and device for medical image Download PDFInfo
- Publication number
- CN111402277A CN111402277A CN202010096775.9A CN202010096775A CN111402277A CN 111402277 A CN111402277 A CN 111402277A CN 202010096775 A CN202010096775 A CN 202010096775A CN 111402277 A CN111402277 A CN 111402277A
- Authority
- CN
- China
- Prior art keywords
- image
- connected domain
- voxel
- segmentation
- outline
- 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.)
- Granted
Links
- 230000011218 segmentation Effects 0.000 title claims abstract description 73
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000004458 analytical method Methods 0.000 claims abstract description 25
- 210000004872 soft tissue Anatomy 0.000 claims description 15
- 238000002591 computed tomography Methods 0.000 claims description 10
- 238000007781 pre-processing Methods 0.000 claims description 9
- 230000002159 abnormal effect Effects 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 3
- 210000003625 skull Anatomy 0.000 claims description 3
- 210000003128 head Anatomy 0.000 description 11
- 210000000613 ear canal Anatomy 0.000 description 8
- 238000010586 diagram Methods 0.000 description 4
- 230000001815 facial effect Effects 0.000 description 2
- 229910010272 inorganic material Inorganic materials 0.000 description 2
- 239000011147 inorganic material Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 241001270131 Agaricus moelleri Species 0.000 description 1
- MXWJVTOOROXGIU-UHFFFAOYSA-N atrazine Chemical compound CCNC1=NC(Cl)=NC(NC(C)C)=N1 MXWJVTOOROXGIU-UHFFFAOYSA-N 0.000 description 1
- 230000003190 augmentative effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 210000001214 frontal sinus Anatomy 0.000 description 1
- 238000002675 image-guided surgery Methods 0.000 description 1
- 238000002595 magnetic resonance imaging Methods 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 210000000214 mouth Anatomy 0.000 description 1
- 210000003928 nasal cavity Anatomy 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 210000003718 sphenoid sinus Anatomy 0.000 description 1
- 238000001356 surgical procedure Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/12—Edge-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
A method and a device for segmenting the outer contour of an object of a medical image can effectively finish the accurate segmentation of the outer contour of the outermost layer of the object aiming at the scanned object with a large number of cavity areas in the scanned object. The method comprises the following steps: (1) normalizing the medical images of different modes, and unifying the gray distribution of the input images; (2) performing binary segmentation on each cross section slice of the image, reserving a maximum background connected domain in a segmentation result, and merging other connected domains; (3) and (3) through connected domain analysis, reserving the maximum background connected domain in each sagittal plane slice of the image, and combining other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
Description
Technical Field
The invention relates to the technical field of medical image processing, in particular to a method for segmenting an object outer contour of a medical image and a device for segmenting the object outer contour of the medical image.
Background
An image-guided surgery navigation system provides important navigation information such as relative position relation of surgical instruments and focus areas for doctors based on medical images such as Computed Tomography (CT) images and Magnetic Resonance Images (MRI) and augmented reality and tracking registration technologies, and assists the doctors in efficiently and safely completing surgery. The outermost contour of the object in the medical image plays an important role in providing visual guide information, intraoperative spatial registration and the like, and is an important component in an operation navigation system.
The voxel value of the scanned object in the medical image is often larger than the voxel value of the air in the background in the medical image, and the scanned object and the air in the background can be distinguished through a single threshold, but most of the scanned objects have a large number of cavity regions similar to the voxel value of the air inside, such as frontal sinuses of the head, sphenoid sinuses and other regions, and some of the cavities are communicated with the outside air, such as nasal cavities, ear canals, oral cavities and other regions in the head, the outer contours formed by the cavities are connected with the outermost layer contour of the object, the outer contours of the cavities are usually retained in the segmentation result obtained through the single threshold segmentation method, and the reconstructed three-dimensional model (as shown in fig. 1. a) cannot meet the requirements of the application scene of the surgical navigation system.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides a method for segmenting the outer contour of an object in a medical image, which can effectively finish the accurate segmentation of the outer contour of the outermost layer of the object aiming at the scanned object with a large number of cavity areas with the voxel values close to that of air.
The technical scheme of the invention is as follows: the object outer contour segmentation method of the medical image comprises the following steps:
(1) normalizing the medical images of different modes, and unifying the gray distribution of the input images;
(2) performing binary segmentation on each cross section slice of the image, reserving a maximum background connected domain in a segmentation result, and merging other connected domains;
(3) and (3) through connected domain analysis, reserving the maximum background connected domain in each sagittal plane slice of the image, and combining other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
The method has the advantages that medical images of different modes are normalized, the gray distribution of input images is unified, each cross section slice of the images is subjected to binary segmentation, the maximum background connected domain in the segmentation result is reserved, other connected domains are merged, the maximum background connected domain in each sagittal plane slice of the images is reserved through connected domain analysis, other connected domains are merged, the segmentation result of the outermost layer contour of the object in the images is obtained, and therefore the scanned object with a large number of cavity regions close to the voxel value of air in the interior can be effectively targeted, and accurate segmentation of the outermost layer contour of the object is completed.
There is also provided an object outline segmentation apparatus for medical imaging, comprising:
the preprocessing module is configured to normalize the medical images of different modalities and unify the gray level distribution of the input images;
the binary segmentation and connected domain analysis module is configured to perform binary segmentation on each cross section slice of the image, reserve the maximum background connected domain in a segmentation result and merge other connected domains;
and the outermost layer contour acquisition module is configured to keep the maximum background connected domain in each sagittal plane slice of the image through connected domain analysis, and merge other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
Drawings
FIG. 1a is a diagram illustrating a single threshold segmentation reconstruction result; fig. 1b, 1c show the difference between CT images.
FIG. 2a is a schematic diagram of the effect of the surface segmentation algorithm, the foreground being shown in black; FIG. 2b is the segmentation result using the Otsu threshold segmentation algorithm; FIG. 2c is a diagram of a foreground label assigned to pixels (headers) greater than a segmentation threshold and a background label assigned to pixels less than the segmentation threshold, followed by connected domain analysis of the background portion of the segmentation result; fig. 2d is a diagram of only the connected component with the largest number of voxels in the connected components, and all other connected components are forced to be the foreground label.
FIG. 3a is a cross-sectional slice of a medical image of an ear canal; 3b is the result of cross-sectional slicing of the ear canal; 3c is the sagittal plane slice connected domain analysis of the ear canal; fig. 3d is the result of processing a sagittal slice of the ear canal.
FIG. 3e is a cross-sectional slice of a medical image within the head; 3f is the result of the processing of the cross-sectional slices in the head; 3g is the sagittal plane slice connected domain analysis result in the head; fig. 3h is the result of processing a sagittal slice within the head.
Fig. 4 is a flowchart of a method for segmenting an outer contour of an object in a medical image according to the present invention.
Detailed Description
As shown in fig. 4, the method for segmenting the outer contour of the object in the medical image includes the following steps:
(1) normalizing the medical images of different modes, and unifying the gray distribution of the input images;
(2) performing binary segmentation on each cross section slice of the image, reserving a maximum background connected domain in a segmentation result, and merging other connected domains;
(3) and (3) through connected domain analysis, reserving the maximum background connected domain in each sagittal plane slice of the image, and combining other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
The method has the advantages that medical images of different modes are normalized, the gray distribution of input images is unified, each cross section slice of the images is subjected to binary segmentation, the maximum background connected domain in the segmentation result is reserved, other connected domains are merged, the maximum background connected domain in each sagittal plane slice of the images is reserved through connected domain analysis, other connected domains are merged, the segmentation result of the outermost layer contour of the object in the images is obtained, and therefore the scanned object with a large number of cavity regions close to the voxel value of air in the interior can be effectively targeted, and accurate segmentation of the outermost layer contour of the object is completed.
Preferably, in the step (1), for the computed tomography image CT, assuming that P (i, j, k) is a gray value of a voxel with an index of (i, j, k) in the image to be segmented, the gray value F [ P (i, j, k) ] of the voxel after preprocessing is obtained by formula (1)
τ is less than the minimum gray value of the scanned object and greater than the gray value of the majority of the noise in the background.
Preferably, τ has a value of-450.
Preferably, in the step (1), for the MRI, histogram statistics is performed on the image before preprocessing to obtain the upper threshold V of the normal voxelMaxMake the voxel value in the image less than VMaxThe ratio of the voxels of (a) to all the voxels of the image is α, and the voxel value is larger than VMaxThe voxel (2) is regarded as an abnormal voxel, and the voxel value is assigned as Vr(ii) a Assuming that P (i, j, k) is the gray value of the voxel with index (i, j, k) in the image to be segmented, obtaining the voxel value G [ P (i, j, k) of the voxel after normalization by formula (2)]：
Preferably, the value of α is 99.9%.
Preferably, in the step (2), each cross section slice of the medical image is sequentially segmented by using an atrazine threshold segmentation algorithm, pixels larger than a segmentation threshold are assigned with foreground labels, and pixels smaller than the segmentation threshold are assigned with background labels; then, performing connected domain analysis on the background part of the segmentation result; only the connected domain with the largest number of voxels in the plurality of connected domains is reserved, and other connected domains are forced to be endowed with foreground labels, as shown in FIGS. 2a-2 d; after all cross sections of the image are sliced, a binary volume data I is obtained0。
In the step (3), after all cross-sectional slices of the image are processed, the outer contours of the two equal cavities still remain in the segmentation result as shown in fig. 3b, but in the binary volume data I0The sagittal section of (a) in which most of the ear canal cavity is separated from the outside to form an independent communicating area, as shown in fig. 3 c; in some medical images, the voxel intensity values of some soft tissues in the human head are small and do not have significant difference from the voxel intensity values of airAnd is generally smaller than the threshold value solved by the Otsu threshold algorithm, if these soft tissues form the same connected domain with the outside, these areas are given background labels, as shown in FIG. 3 f; if these soft tissues are distributed in the skull, the volume data I will be0Becomes an independent connected domain in the sagittal slice as shown in FIG. 3 g; for both cases, the outermost layer profile of the object can be obtained by connected domain analysis, as shown in fig. 3d and 3 h. Preferably, in the step (3), after all cross-sectional slices of the image are processed, the volume data I are sequentially processed0And performing connected domain analysis on the background regions of all the sagittal plane slices, reserving the connected domain with the largest number of voxels, and assigning other connected domains as foreground labels to obtain the outermost layer contour surface of the object.
Preferably, for medical images of the head, processing is performed on cross-sectional slices as well as sagittal slices of the images; for the images of other objects, the direction and sequence of processing are selected according to specific situations.
It will be understood by those skilled in the art that all or part of the steps in the method of the above embodiments may be implemented by hardware instructions related to a program, the program may be stored in a computer-readable storage medium, and when executed, the program includes the steps of the method of the above embodiments, and the storage medium may be: ROM/RAM, magnetic disks, optical disks, memory cards, and the like. Therefore, corresponding to the method of the present invention, the present invention also includes an apparatus for segmenting the outer contour of the object in medical images, which is generally expressed in the form of functional modules corresponding to the steps of the method. The device includes:
the preprocessing module is configured to normalize the medical images of different modalities and unify the gray level distribution of the input images;
the binary segmentation and connected domain analysis module is configured to perform binary segmentation on each cross section slice of the image, reserve the maximum background connected domain in a segmentation result and merge other connected domains;
and the outermost layer contour acquisition module is configured to keep the maximum background connected domain in each sagittal plane slice of the image through connected domain analysis, and merge other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image. The present invention is described in more detail below.
Specific examples of the present invention are described in more detail below. The object outline segmentation method of the medical image comprises the following steps:
(1) pretreatment of
The gray distribution ranges of the images in different modes have larger difference, and the medical images are respectively normalized according to different modes.
The gray scale distribution range of Computed Tomography (CT) is relatively fixed, but the difference between CT images may also occur due to different types and models of scanning machines, for example, image information collected by some machines may be distributed in the whole image space, and some may be distributed only inside the cylinder, as shown in fig. 1b and 1 c. To eliminate the influence of the difference between images and background noise on the segmentation result, preprocessing is performed using the following simple operation. Let P (i, j, k) be the gray value of the voxel with index (i, j, k) in the image to be segmented, and F [ P (i, j, k) ] be the gray value of the voxel after preprocessing.
Wherein tau is selected according to the practical application scene, tau should be smaller than the minimum gray value of the scanned object and larger than the gray value of most noise in the background in order to eliminate the image caused by the segmentation of the background noise (inorganic material such as cloth liner) as much as possible, and the default value of tau is-450.
Unlike CT images, Magnetic Resonance Imaging (MRI) images by detecting the spin of water molecules, and since inorganic materials contain substantially no water, the MRI images have less background noise, but the grayscale distribution range of the MRI images is not fixed, so it is necessary to perform line mapping on all images to unify the distribution range of voxel intensity values of all images to [0, V [ ]r]. In order to eliminate the influence of abnormal large voxels in the image on the normalization operation, the image is subjected to histogram before preprocessingCarrying out image statistics to obtain an upper limit threshold V of normal voxelsMaxThe voxel value in the image is less than VMaxThe ratio of the voxels in (a) to all voxels in (b) the image is α (α is 99.9% by default), and the voxel value is greater than VMaxThe voxel (2) is regarded as an abnormal voxel, and the voxel value is assigned as Vr. Let P (i, j, k) be the gray value of the voxel with index (i, j, k) in the image to be segmented, G [ P (i, j, k)]To normalize the voxel value of this voxel, then:
(2) binary segmentation based on connected component analysis
The scanning process of the medical image is standardized, and in general, after the coordinate system of the medical image is normalized according to the scanning information of the medical image data, the image slices in three dimensions of the medical image can respectively reflect the anatomical features of the cross section, the sagittal plane and the coronal plane of the patient. After the medical image is subjected to threshold segmentation, certain cavities in the medical image, which are communicated with the outside, and the outside air form the same communicated domain, but slices of the internal cavities in most fault planes of the image are not communicated with the outside, the image is processed in a layering mode, the largest background communicated domain in each layer of image is reserved, the cavities can be removed, and the segmentation result of the outermost layer contour of the object is obtained.
Each slice of the cross section of the medical image is first processed. Because the gray scales of the background area and the scanned object have obvious difference, the algorithm selects an Otsu threshold segmentation algorithm to segment each cross section slice of the medical image in sequence, the segmentation result is shown as figure 2b, pixels (heads) larger than the segmentation threshold are endowed with foreground labels, pixels smaller than the segmentation threshold are endowed with background labels, then connected domain analysis is carried out on the background part of the segmentation result, as shown in figure 2c, only the connected domain with the largest number of voxels in a plurality of connected domains is reserved, other connected domains are all forcibly endowed with foreground labels, and the obtained result is shown as figure 2 d. After all cross sections of the image are sliced, a binary volume data I is obtained0。
Since most of the cavities communicating with the outside are still connected to the outside in the cross-sectional slices of the image, such as the ear canal, after all the cross-sectional slices of the image are processed, the outline of the cavities still remains in the segmentation result, as shown in fig. 3b, but in the binary volume data I0The sagittal section of (3 c) shows that most of the ear canal cavity is separated from the outside to form an independent communicating area. In some medical images, the voxel intensity values of some soft tissues in the human head are small, no significant difference exists between the voxel intensity values of the soft tissues and the voxel intensity values of the air, and the voxel intensity values are generally smaller than the threshold value solved by the Otsu threshold algorithm, if the soft tissues and the outside form the same connected domain, the soft tissues are assigned with background labels (as shown in fig. 3 f), but since the soft tissues are generally distributed in the skull, the volume data I0Will become an independent connected domain in the sagittal plane slice (as shown in fig. 3 g), and the outermost contour of the object can be obtained by the connected domain analysis (as shown in fig. 3 h).
After processing all cross section slices of the image, the algorithm sequentially compares the volume data I0And (3) performing connected domain analysis on the background regions of all sagittal plane slices, as shown in fig. 3c and 3g, reserving the connected domain with the largest number of voxels, and assigning other connected domains as foreground labels to obtain the final target segmentation result of the algorithm, as shown in fig. 3d and 3h, namely the outermost layer contour surface of the object.
Due to the concave and convex parts of the face organs and the skin folds, internal connected domains similar to the internal connected domains may exist in the binary segmentation result of the coronal section of the image, but the internal connected domains just reflect facial feature information, and if the coronal section of the image is processed by using the algorithm, the facial information of the patient is damaged, so that the medical image of the head can be processed only on the transversal section and the sagittal section of the image. For images of other objects, the direction and order of processing can be selected according to specific situations.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the present invention in any way, and all simple modifications, equivalent variations and modifications made to the above embodiment according to the technical spirit of the present invention still belong to the protection scope of the technical solution of the present invention.
Claims (10)
1. An object outline segmentation method of medical images is characterized in that: which comprises the following steps:
(1) normalizing the medical images of different modes, and unifying the gray distribution of the input images;
(2) performing binary segmentation on each cross section slice of the image, reserving a maximum background connected domain in a segmentation result, and merging other connected domains;
(3) and (3) through connected domain analysis, reserving the maximum background connected domain in each sagittal plane slice of the image, and combining other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
2. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: in the step (1), for the CT, assuming P (i, j, k) as the gray value of the voxel with index (i, j, k) in the image to be segmented, the preprocessed gray value F [ P (i, j, k) ]ofthe voxel is obtained by formula (1)
τ is less than the minimum gray value of the scanned object and greater than the gray value of the majority of the noise in the background.
3. The method for segmenting the outline of an object in a medical image according to claim 2, wherein: in the CT preprocessing method for the computed tomography image in the step (1), the value of tau is-450.
4. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: in the step (1), before the pretreatment, the MRIHistogram statistics is carried out on the image to obtain an upper limit threshold V of a normal voxelMaxMake the voxel value in the image less than VMaxThe ratio of the voxels of (a) to all the voxels of the image is α, and the voxel value is larger than VMaxThe voxel (2) is regarded as an abnormal voxel, and the voxel value is assigned as Vr(ii) a Assuming that P (i, j, k) is the gray value of the voxel with index (i, j, k) in the image to be segmented, obtaining the voxel value G [ P (i, j, k) of the voxel after normalization by formula (2)]：
5. The method for segmenting the outline of a subject according to claim 4, wherein the value of α in the MRI preprocessing method for the magnetic resonance image in step (1) is 99.9%.
6. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: in the step (2), each cross section slice of the medical image is sequentially segmented by using an Otsu threshold segmentation algorithm, pixels larger than a segmentation threshold are assigned with foreground labels, and pixels smaller than the segmentation threshold are assigned with background labels; then, performing connected domain analysis on the background part of the segmentation result; only the connected domain with the largest number of voxels in the plurality of connected domains is reserved, and other connected domains are forced to be endowed with foreground labels; after all cross sections of the image are sliced, a binary volume data I is obtained0。
7. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: in the step (3), after all cross-sectional slices of the image are processed, the outline of the cavities still remains in the segmentation result, but the binary volume data I0In the sagittal section, most of the auditory canal cavity is separated from the outside to form an independent communicating area; in some medical images, some soft tissue in the human headThe voxel intensity value of the soft tissue is smaller, no obvious difference exists between the soft tissue and the voxel intensity value of air, and the soft tissue is usually smaller than a threshold value solved by an Otsu threshold algorithm, and if the soft tissue and the outside form the same connected domain, a background label is given to the soft tissue; if the soft tissues are distributed in the skull, the volume data I will be0Becomes an independent connected domain in the sagittal plane slice, and obtains the outermost layer outline of the object through the analysis of the connected domain.
8. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: in the step (3), after all cross sections of the image are processed, the volume data I are sequentially subjected to volume data I0And performing connected domain analysis on the background regions of all the sagittal plane slices, reserving the connected domain with the largest number of voxels, and assigning other connected domains as foreground labels to obtain the outermost layer contour surface of the object.
9. The method for segmenting the outline of an object in a medical image according to claim 1, wherein: processing the cross section slice and the sagittal plane slice of the image aiming at the medical image of the head; for the images of other objects, the direction and sequence of processing are selected according to specific situations.
10. An object outline segmentation device of medical image is characterized in that: it includes:
the preprocessing module is configured to normalize the medical images of different modalities and unify the gray level distribution of the input images;
the binary segmentation and connected domain analysis module is configured to perform binary segmentation on each cross section slice of the image, reserve the maximum background connected domain in a segmentation result and merge other connected domains;
and the outermost layer contour acquisition module is configured to keep the maximum background connected domain in each sagittal plane slice of the image through connected domain analysis, and merge other connected domains to obtain the segmentation result of the outermost layer contour of the object in the image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010096775.9A CN111402277B (en) | 2020-02-17 | 2020-02-17 | Object outline segmentation method and device for medical image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010096775.9A CN111402277B (en) | 2020-02-17 | 2020-02-17 | Object outline segmentation method and device for medical image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111402277A true CN111402277A (en) | 2020-07-10 |
| CN111402277B CN111402277B (en) | 2023-11-14 |
Family
ID=71434260
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010096775.9A Active CN111402277B (en) | 2020-02-17 | 2020-02-17 | Object outline segmentation method and device for medical image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111402277B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112967192A (en) * | 2021-02-22 | 2021-06-15 | 艾瑞迈迪医疗科技（北京）有限公司 | Depth perception enhancement method and device based on 2D/3D vascular fusion |
| CN113643263A (en) * | 2021-08-18 | 2021-11-12 | 北京理工大学 | Identification method and system for upper limb bone positioning and forearm bone fusion deformity |
| CN114298927A (en) * | 2021-12-16 | 2022-04-08 | 北京朗视仪器股份有限公司 | CBCT image denoising method and device, storage medium and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657984A (en) * | 2015-01-28 | 2015-05-27 | 复旦大学 | Automatic extraction method of three-dimensional breast full-volume image regions of interest |
| CN107610145A (en) * | 2017-07-26 | 2018-01-19 | 同济大学 | A kind of automatic pancreas dividing method based on adaptive threshold and template matches |
| US20190066303A1 (en) * | 2017-08-31 | 2019-02-28 | Council Of Scientific & Industrial Research | Method And System For Automatic Volumetric-Segmentation Of Human Upper Respiratory Tract |
| CN109583444A (en) * | 2018-11-22 | 2019-04-05 | 博志生物科技有限公司 | Hole region localization method, device and computer readable storage medium |
| CN109961446A (en) * | 2019-03-27 | 2019-07-02 | 深圳视见医疗科技有限公司 | CT/MR three-dimensional image segmentation processing method, device, equipment and medium |
-
2020
- 2020-02-17 CN CN202010096775.9A patent/CN111402277B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104657984A (en) * | 2015-01-28 | 2015-05-27 | 复旦大学 | Automatic extraction method of three-dimensional breast full-volume image regions of interest |
| CN107610145A (en) * | 2017-07-26 | 2018-01-19 | 同济大学 | A kind of automatic pancreas dividing method based on adaptive threshold and template matches |
| US20190066303A1 (en) * | 2017-08-31 | 2019-02-28 | Council Of Scientific & Industrial Research | Method And System For Automatic Volumetric-Segmentation Of Human Upper Respiratory Tract |
| CN109583444A (en) * | 2018-11-22 | 2019-04-05 | 博志生物科技有限公司 | Hole region localization method, device and computer readable storage medium |
| CN109961446A (en) * | 2019-03-27 | 2019-07-02 | 深圳视见医疗科技有限公司 | CT/MR three-dimensional image segmentation processing method, device, equipment and medium |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112967192A (en) * | 2021-02-22 | 2021-06-15 | 艾瑞迈迪医疗科技（北京）有限公司 | Depth perception enhancement method and device based on 2D/3D vascular fusion |
| CN113643263A (en) * | 2021-08-18 | 2021-11-12 | 北京理工大学 | Identification method and system for upper limb bone positioning and forearm bone fusion deformity |
| CN114298927A (en) * | 2021-12-16 | 2022-04-08 | 北京朗视仪器股份有限公司 | CBCT image denoising method and device, storage medium and electronic equipment |
| CN114298927B (en) * | 2021-12-16 | 2022-10-04 | 北京朗视仪器股份有限公司 | CBCT image denoising method and device, storage medium and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111402277B (en) | 2023-11-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111402277B (en) | Object outline segmentation method and device for medical image | |
| Kapur et al. | Segmentation of brain tissue from magnetic resonance images | |
| US6249594B1 (en) | Autosegmentation/autocontouring system and method | |
| JP6034299B2 (en) | System for estimating penumbra size and method of operation thereof | |
| JP2011514190A (en) | Method and system for segmentation and modeling of anatomical structures in images | |
| CN110120048B (en) | Three-dimensional brain tumor image segmentation method combining improved U-Net and CMF | |
| EP2715663B1 (en) | Apparatus for generating assignments between image regions of an image and element classes | |
| CN106997594B (en) | Method and device for positioning eye tissue | |
| CN112184720B (en) | Method and system for segmenting internal rectus muscle and optic nerve of CT image | |
| CN114627067A (en) | Wound area measurement and auxiliary diagnosis and treatment method based on image processing | |
| Kale et al. | Automatic segmentation of human facial tissue by MRI–CT fusion: A feasibility study | |
| Liu et al. | Automatic segmentation of the human brain ventricles from MR images by knowledge-based region growing and trimming | |
| Shen et al. | Localized priors for the precise segmentation of individual vertebras from ct volume data | |
| Dogdas et al. | Segmentation of the skull in 3D human MR images using mathematical morphology | |
| Ukil et al. | Smoothing lung segmentation surfaces in 3D X-ray CT images using anatomic guidance | |
| Hernandez et al. | Three-dimensional segmentation of brain aneurysms in CTA using non-parametric region-based information and implicit deformable models: Method and evaluation | |
| CN115841472A (en) | Method, device, equipment and storage medium for identifying high-density characteristics of middle cerebral artery | |
| EP1080449B1 (en) | Autosegmentation / autocontouring system | |
| Roy et al. | MDL-IWS: multi-view deep learning with iterative watershed for pulmonary fissure segmentation | |
| CN113222886B (en) | Jugular fossa and sigmoid sinus groove positioning method and intelligent temporal bone image processing system | |
| Lee et al. | Unsupervised abnormalities extraction and brain segmentation | |
| CN114463304B (en) | Method for segmenting and labeling brain atlas and operation navigation system | |
| 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 | |
| Paccini et al. | Mapping Grey-Levels on 3D Segmented Anatomical districts. |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |