CN104809723B - The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure - Google Patents
The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure Download PDFInfo
- Publication number
- CN104809723B CN104809723B CN201510173307.6A CN201510173307A CN104809723B CN 104809723 B CN104809723 B CN 104809723B CN 201510173307 A CN201510173307 A CN 201510173307A CN 104809723 B CN104809723 B CN 104809723B
- Authority
- CN
- China
- Prior art keywords
- liver
- image
- algorithm
- super voxel
- area
- 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.)
- Expired - Fee Related
Links
Abstract
The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure, by analyzing volume data histogram, adaptively strengthens picture contrast.Liver initial profile segmentation is successively carried out using adaptive threshold and morphological method, chooses maximum liver section, calculates and extracts liver area-of-interest.In maximum liver section according to liver initial profile selected seed point, prospect background color is modeled using gauss hybrid models.Super voxel is generated using SLIC clustering algorithms to the liver area-of-interest after enhancing contrast, undirected weighted graph is constructed by summit of super voxel, cuts algorithm using figure and figure is cut.Finally segmentation result is post-processed using morphology operations etc..Quick, the accurate automatic segmentation of liver in three-dimensional abdominal CT images can be achieved in the present invention.
Description
Technical field
The present invention relates to field of medical image processing, specifically, is related to the three-dimensional liver that algorithm is cut based on super voxel and figure
Dirty CT image automatic segmentation methods.
Background technology
Primary carcinoma of liver is one of malignant tumour common in the world, has very high morbidity and mortality.Computer
Tomoscan (ComputedTomography, CT) imaging technique due to anatomic information is accurate, high resolution, sweep time
Short, the advantages that popularity rate is high, it is widely used in the Clinics and Practices of liver cancer.Accurate liver three-dimensional segmentation is area of computer aided
A basic work in diagnosis, it is the important prerequisite of three-dimensional visualization, quantitative analysis, surgery planning etc..Face at present
The segmentation of liver is typically rule of thumb had been manually done by doctor on bed, this not only time and effort consuming, and the degree of accuracy varies with each individual.Therefore
The liver of efficient stable is partitioned into study hotspot automatically, and, for the working strength of mitigation doctor, improving diagnosis speed has for this
Greatly help.
The automatic segmentation for realizing liver is a challenging research work, and its difficult point is：Liver structure is complicated,
Own form is changeable, having differences property between individual, because liver organization gray scale is uneven caused by tumour, hepatic sclerosis, with adjoining
The obscurity boundary of the histoorgans such as diaphram, heart, kidney.In recent decades, both at home and abroad in terms of CT image for liver segmentation
Many methods are proposed, are mainly included：Region growing, active contour, level set, figure cut, clustered, statistical shape model and probability graph
Spectrum etc..Algorithm of region growing have the advantages that it is quick, easily realize, but easily cause segmentation result when liver organization gray scale is uneven
It is inaccurate.Partitioning algorithm based on active contour and level set need to provide initial profile and calculating complexity, splitting speed are slower.Base
Although can obtain accurate segmentation result in the partitioning algorithm of model, the generation of probability collection of illustrative plates or statistical shape model needs
Substantial amounts of training image and corresponding artificial segmentation standard are wanted, and segmentation result may be caused when handling non-standard shapes liver
It is inaccurate.Figure cuts algorithm and is widely used in medical image segmentation because can obtain globally optimal solution, however directly using voxel as
Unit structure figure, which carries out cutting, can cause too high amount of calculation and can not obtain satisfied segmentation result.Accurately, fast and automatically change
Segmentation be liver segmentation target.
The content of the invention
The shortcomings that it is an object of the invention to overcome prior art and deficiency, propose that a kind of simple linear iteration that is based on clusters
(Simple Linear Iterative Clustering, SLIC) and figure cut the three-dimensional CT image for liver of the algorithm side of segmentation automatically
Method.This method can adaptively strengthen picture contrast, automatically extract liver area-of-interest and generate super voxel, and with super voxel
Undirected weighted graph is built for elementary cell, computation complexity and amount of calculation is greatly reduced, improves treatment effeciency.And can reaches simultaneously
Automatic selected seed point uses restraint to segmentation, and establishes liver and the color model of background area using gauss hybrid models,
Avoid influence of the man-machine interactively selected seed point to algorithm robustness.This method cuts algorithm to liver area-of-interest using figure
The super undirected weighted graph of voxel is cut, and quick, accurate and automation liver segmentation can be realized, so as to mitigate working doctor
Amount, auxiliary is provided to medical diagnosis.
The present invention improve it is a kind of voxel is surpassed based on SLIC and figure cuts the three-dimensional CT image for liver automatic division method of algorithm,
Comprise the following steps：
1.1. histogram analysis are carried out to an abdominal CT images I of input, so as to adaptively strengthen picture contrast, obtained
CT images I ' to after enhancing contrast；
1.2. liver initial profile segmentation is successively carried out to image I using adaptive threshold and morphological method, chosen most
Big liver section lmax, calculate and to image I ' extractions liver area-of-interest I ' after enhancingROI；
1.3. in maximum liver section lmaxOn liver seed point S chosen according to liver initial profileFWith background seed point SB,
The gray probability that liver and background area are calculated using the method based on gauss hybrid models is distributed, and obtains gauss hybrid models Pfg
And Pbkg；
1.4. to I 'ROIOver-segmentation, which is carried out, using SLIC algorithms generates super voxel；
1.5. undirected weighted graph G is built by summit of super voxel, cuts algorithm using figure and figure G is cut, obtain liver area
Regional partition result bianry image Imask；
1.6. using morphology opening operation, medium filtering to ImaskPost-process, obtain smooth liver segmentation results.
The step 1.1 includes,
1.1.1. peak value number and peak C T values in image histogram are analyzed, is height if in the presence of 2 obvious peak values
Contrast image Ihigh, it is soft image I if 1 peak value is only existedlow；
1.1.2. threshold range [the I of adaptive contrast stretching is calculatedmin,Imax], for high-contrast image IhighHave
Imin=(peak1+peak2)/2-60, Imax=peak2+250；For low comparison diagram image IlowThere is Imin=peak1-60, Imax
=peak1+250, wherein peak1, peak2 represent peak C T values；
1.1.3. contrast enhancing, I '=255 (I-I aremin)/(Imax-Imin)。
The step 1.2 includes,
1.2.1. adaptive threshold [peak1,400Hu] method extraction liver Probability Area is utilized for each layer of section l
lpossible；
1.2.2. to lpossibleSuccessively using morphological erosion, the maximum subregion of selection, holes filling, morphological dilations side
Method, obtain the initial profile l of liver arealiver；
1.2.3. l is chosenliverThe maximum conduct maximum liver section l of middle liver area areamax；
1.2.4. according to every layer of lliverTwo-dimentional liver area bounding box be calculated that to surround whole three-dimensional liver sense emerging
The minimum bounding box in interesting region, the middle extraction liver area-of-interest I ' of image I ' after enhancingROI。
The step 1.3 includes,
1.3.1. for maximum liver section lmax, in liver initial profile lliverIt is internal to choose liver kind at regular intervals
Sub- point SF, lliverBounding box chooses background seed point S with exterior domain at regular intervalsB；
1.3.2. according to seed point SF, SBCalculate gauss hybrid models Pfg, Pbkg,
Wherein, IpFor gray value corresponding to voxel, k is the number of components of gauss hybrid models, ω, μ and σ2Respectively Gauss point
Weight, average and the variance of amount.
The step 1.4 includes,
1.4.1. SLIC clustering algorithm formula, N=kS are provided3, wherein N is I 'ROIThe total number of middle voxel, S are super voxel
Spacing, k are the super voxel number ultimately produced；
1.4.2. clustered according to distance measure D,
Wherein, dc, dsRespectively color and distance spatially, m is weight, sets m=20, S=3, each super voxel
Gray value is set to the gray average in the super voxel areas.
In the step 1.5, provide figure and cut algorithm energy equation,R(lp
=0)=- log (Pfg(Ip)), R (lp=1)=- log (Pbkg(Ip)), B (lp,lq)=δ (lp,lq)/((Ip-Iq)2+1).Wherein R
(lp) and B (lp,lq) it is respectively area item and border item,
The present invention compared with prior art, has the following advantages that and beneficial effect：
1. the inventive method automatization level when splitting CT image for liver is high, it is complete, sufficient automatic segmentation, keeps away
Having exempted from most of partitioning algorithms needs influence caused by man-machine interactively to algorithm robustness.
2. the present invention effectively reduces the computation complexity that figure cuts algorithm, real-time is improved, processing speed is than existing
Three-dimensional fast an order of magnitude of liver automatic division method.
3. 30 clinical abdominal CT images that inventive algorithm is provided using MICCAI 2007Workshop are verified,
For contrast it is relatively low, tumour be present when CT images can carry out quickly and accurately automatically split.
Brief description of the drawings
Fig. 1：The flow chart of the inventive method；
Fig. 2：CT image histograms example in the inventive method, wherein：(a) high-contrast image histogram, (b) low contrast
Spend image histogram；
Fig. 3：Maximum liver section is chosen in the inventive method and calculates the flow chart of liver region of interest border frame；
Fig. 4：A certain sectioning image in original CT sequence to be split；
Fig. 5：The design sketch after Adaptive contrast enhancement is done to original CT image；
Fig. 6：Liver initial profile is done to the image after enhancing contrast using adaptive threshold and morphological method to split
Bianry image afterwards；
Fig. 7：To the design sketch after the image zooming-out liver area-of-interest after enhancing contrast；
Fig. 8：The super voxel of over-segmentation generation is carried out using SLIC algorithms to the liver region of interest area image of extraction；
Fig. 9：Final liver segmentation results bianry image；
Figure 10：To the segmentation effect of the uneven image of liver organization gray scale, wherein (a) is cross section, (b) is coronal-plane,
(c) it is sagittal plane.
Figure 11：To the segmentation effect with non-standard liver shape image, wherein (a) is cross section, (b) is coronal-plane,
(c) it is sagittal plane.
Embodiment
Extraction process is specifically described with reference to accompanying drawing and concrete instance.Institute comes from MICCAI using view data
Belly Enhanced CT image in 2007Workshop databases.The mean size of every CT image is 512*512*208 pictures
Element, average mark resolution are 0.68*0.68*1.6 millimeters.
Flow chart such as Fig. 1 institutes of the CT image for liver automatic division method that algorithm is cut based on super voxel and figure of the present invention
Show, comprise the following steps：
Step 1, to the abdominal CT images I progress histogram analysis (as shown in Figure 4) of input, image is adaptively strengthened
Contrast, obtain strengthening the CT images I ' (as shown in Figure 5) after contrast.Specific implementation step is as follows：
1.1. peak value number in image histogram is analyzed, is high-contrast image if in the presence of 2 obvious peak values
Ihigh(as shown in Figure 2 a) it is, soft image I if 1 peak value is only existedlow(as shown in Figure 2 b) each peak C T, is recorded
Value peak1, peak2；
1.2. threshold range [the I of adaptive contrast stretching is calculatedmin,Imax],
For high-contrast image IhighHave：
Imin=(peak1+peak2)/2-60
Imax=peak2+250
For soft image IlowHave：
Imin=peak1-60
Imax=peak1+250
Linear contrast's stretching is done to original image, CT values are less than I in imageminSet to 0, CT values are more than ImaxBe set to 255,
CT values are in [Imin,Imax] in the range of have：
Step 2, liver initial profile segmentation is successively carried out to image I using adaptive threshold and morphological method, chosen
Maximum liver section lmax, calculate three-dimensional liver region of interest ROI (flow chart is as shown in Figure 3).Specific implementation step is as follows：
2.1. adaptive threshold method extraction liver Probability Area l is utilized for each layer of section lpossible：
1) threshold value is set as [peak1,400Hu], and voxel CT values are less than peak1 or setting to 0 more than 400Hu, and CT values are in threshold
255 are set in the range of value, is target area.The liver Probability Area extracted further comprises some gray values in addition to liver
Similar tissue and organ；
2) the circular mask that actionradius is 7 is to lpossibleDo morphological erosion；
3) retain the maximum subregion on the left of image, remove other subregions；
4) cavity filled up inside region, the circular mask that actionradius is 2 do morphological dilations to image, finally given
Liver initial profile (as shown in Figure 6).
2.2. l is chosenliverThe maximum conduct maximum liver section l of middle liver area areamax；
2.3. according to every layer of lliverTwo-dimentional liver area bounding box be calculated that to surround whole three-dimensional liver interested
The minimum bounding box in region；
2.4. the middle extraction liver area-of-interest I ' of image I ' after enhancingROI(as shown in Figure 7).
Step 3, in maximum liver section lmaxUpper selection target liver seed point SFWith background seed point SB, using based on height
The method of this mixed model calculates the gray probability distribution of liver and background area, obtains gauss hybrid models PfgAnd Pbkg.Specifically
Implementation steps are as follows：
3.1. in liver initial profile lliverLiver seed point S is chosen by the interval of 30 pixels in insideF.In two-dimentional liver area
Outside the bounding box of domain the non-zero background seed point S of gray value is chosen by the interval of 20 pixelsB；
3.2. the gray probability that liver and background are calculated according to the gray value of seed point set is distributed, and is expressed as Gauss and mixes
Matched moulds type Pfg, Pbkg.Gauss hybrid models formula is：
Set the number of components k=5 of gauss hybrid models.Each Gaussian component is expressed as：
Wherein, ω, μ and σ2The respectively weight of Gaussian component, average and variance,
N represents the number of seed point.
Step 4, to the liver area-of-interest I ' after enhancing contrastROIIt is super that over-segmentation generation is carried out using SLIC algorithms
Voxel (as shown in Figure 8).Specific implementation step is as follows：
4.1.N=kS3, set super voxel interval S=3 of generation.N is I 'ROIThe total number of middle voxel.Then in I 'ROIIn
S generates k initial cluster center at regular intervals；
4.2. centered on each cluster centre point, it is limited in the spatial dimension of 2S*2S*2S sizes and scans for.It is right
In each voxel in search space, D is estimated according to minimum range and clustered,
Wherein, dc, dsRespectively color Ii,IjWith the distance on space [x, y, z].M is weight.Set m=20, S=3.
Each super voxel is a small cluster areas, and its gray value is set to the average of all voxel gray values in the cluster areas.
Step 5, undirected weighted graph G is built by summit V of super voxel.Algorithm is cut using figure to cut figure G, obtains liver
Region segmentation result bianry image Imask.Specific implementation step is as follows：
5.1. segmentation problem is converted into and minimizes energy equation E (L),
α, β are weighting parameters.For each super voxel p, there is lp∈ { 0,1 }, wherein label 0 represent background, and label 1 represents
Prospect (liver).R(lp) it is area item,
R(lp=0)=- log (Pfg(Ip)), R (lp=1)=- log (Pbkg(Ip))
B(lp,lq) it is border item,
5.2. undirected weighted graph G=(V, N) is built by summit V of super voxel, N represents the nonoriented edge collection of figure, adds 2 ends
End points S, T.The weight on each side is as shown in table 1.Set S corresponding labels 0 (background), T corresponding labels 1 (target liver)；
The weight on each side of the figure of table 1
5.3. cut algorithm using figure to cut the weighted graph of structure, obtain liver area segmentation result bianry image
Imask。
Step 6, segmentation result bianry image is smoothed successively using morphology opening operation, median filter.
Obtain final three-dimensional liver segmentation results (as shown in Figure 9).
For liver organization gray scale it is uneven, the CT images of tumour be present, this method can carry out quickly and accurately automatic
Segmentation, segmentation result is as shown in Figure 10,11.
Claims (6)
1. the three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure, it is characterised in that：This method includes
Following steps,
1.1. histogram analysis are carried out to an abdominal CT images I of input, so as to adaptively strengthen picture contrast, increased
CT images I ' after strong contrast；
1.2. liver initial profile segmentation is successively carried out to image I using adaptive threshold and morphological method, chooses maximum liver
Dirty section lmax, calculate and to image I ' extractions liver area-of-interest I ' after enhancingROI；
1.3. in maximum liver section lmaxOn liver seed point S chosen according to liver initial profileFWith background seed point SB, use
Method based on gauss hybrid models calculates the gray probability distribution of liver and background area, obtains gauss hybrid models model Pfg
And Pbkg；
1.4. to I 'ROIOver-segmentation, which is carried out, using SLIC algorithms generates super voxel；
1.5. undirected weighted graph G is built by summit of super voxel, cuts algorithm using figure and figure G is cut, obtain liver area point
Cut result bianry image Imask；
1.6. using morphology opening operation, medium filtering to ImaskPost-process, obtain smooth liver segmentation results.
2. the three-dimensional CT image for liver automatic division method according to claim 1 that algorithm is cut based on super voxel and figure, its
It is characterised by：Described step 1.1 includes,
1.1.1. peak value number and peak C T values in image histogram are analyzed, is high contrast if in the presence of 2 obvious peak values
Spend image Ihigh, it is soft image I if 1 peak value is only existedlow；
1.1.2. threshold range [the I of adaptive contrast stretching is calculatedmin,Imax], for high-contrast image IhighThere is Imin=
(peak1+peak2)/2-60, Imax=peak2+250；For low comparison diagram image IlowThere is Imin=peak1-60, Imax=
Peak1+250, wherein peak1, peak2 represent peak C T values；
1.1.3. contrast enhancing, I '=255 (I-I aremin)/(Imax-Imin)。
3. the three-dimensional CT image for liver automatic division method according to claim 1 that algorithm is cut based on super voxel and figure, its
It is characterised by：Described step 1.2 includes,
1.2.1. adaptive threshold [peak1,400Hu] method extraction liver Probability Area is utilized for each layer of section l
lpossible；
1.2.2. to lpossibleMorphological erosion is used successively, retains maximum subregion, holes filling, morphological dilations method,
Obtain the initial profile l of liver arealiver；
1.2.3. l is chosenliverThe maximum conduct maximum liver section l of middle liver area areamax；
1.2.4. according to every layer of lliverTwo-dimentional liver area bounding box be calculated and can surround whole three-dimensional liver region of interest
The minimum bounding box in domain, the middle extraction liver area-of-interest I ' of image I ' after enhancingROI。
4. the three-dimensional CT image for liver automatic division method according to claim 1 that algorithm is cut based on super voxel and figure, its
It is characterised by：Described step 1.3 includes,
1.3.1. for maximum liver section lmax, in liver initial profile lliverIt is internal to choose liver seed point at regular intervals
SF, lliverBounding box chooses background seed point S with exterior domain at regular intervalsB；
1.3.2. according to seed point SF, SBCalculate gauss hybrid models Pfg, Pbkg,
Wherein, IpFor gray value corresponding to voxel, k is the number of components of gauss hybrid models, ω, μ and σ2Respectively Gaussian component
Weight, average and variance.
5. the three-dimensional CT image for liver automatic division method according to claim 1 that algorithm is cut based on super voxel and figure, its
It is characterised by：Described step 1.4 includes,
1.4.1. SLIC clustering algorithm formula, N=kS are provided3, wherein N is I 'ROIThe total number of middle voxel, S are super voxel spacing,
K is the super voxel number ultimately produced；
1.4.2. clustered according to distance measure D,
Wherein, dc, dsRespectively color and distance spatially, m is weight, sets m=20, S=3, the gray scale of each super voxel
Value is set to the gray average in the super voxel areas.
6. the three-dimensional CT image for liver automatic division method according to claim 1 that algorithm is cut based on super voxel and figure, its
It is characterised by：In described step 1.5, provide figure and cut algorithm energy equation,R
(lp=0)=- log (Pfg(Ip)), R (lp=1)=- log (Pbkg(Ip)), B (lp,lq)=δ (lp,lq)/((Ip-Iq)2+ 1), its
In, R (lp) and B (lp,lq) it is respectively area item and border item,
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510173307.6A CN104809723B (en) | 2015-04-13 | 2015-04-13 | The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510173307.6A CN104809723B (en) | 2015-04-13 | 2015-04-13 | The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104809723A CN104809723A (en) | 2015-07-29 |
| CN104809723B true CN104809723B (en) | 2018-01-19 |
Family
ID=53694528
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510173307.6A Expired - Fee Related CN104809723B (en) | 2015-04-13 | 2015-04-13 | The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104809723B (en) |
Families Citing this family (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139454B (en) * | 2015-08-06 | 2018-01-19 | 北京工业大学 | The extraction method of liver three dimensional area of interest in a kind of three-dimensional CT image |
| CN105286988A (en) * | 2015-10-12 | 2016-02-03 | 北京工业大学 | CT image-guided liver tumor thermal ablation needle location and navigation system |
| WO2017092615A1 (en) | 2015-11-30 | 2017-06-08 | 上海联影医疗科技有限公司 | Computer aided diagnosis system and method |
| CN105574859B (en) * | 2015-12-14 | 2018-08-21 | 中国科学院深圳先进技术研究院 | A kind of liver neoplasm dividing method and device based on CT images |
| CN106898044B (en) * | 2017-02-28 | 2020-08-04 | 成都金盘电子科大多媒体技术有限公司 | Organ splitting and operating method and system based on medical images and by utilizing VR technology |
| CN107146228B (en) * | 2017-03-22 | 2019-10-08 | 东南大学 | A kind of super voxel generation method of brain magnetic resonance image based on priori knowledge |
| CN107016681B (en) * | 2017-03-29 | 2023-08-25 | 浙江师范大学 | Brain MRI tumor segmentation method based on full convolution network |
| CN107038705B (en) * | 2017-05-04 | 2020-02-14 | 季鑫 | Retinal image bleeding area segmentation method and device and computing equipment |
| CN107516314B (en) * | 2017-08-25 | 2020-03-27 | 山东大学 | Medical image hyper-voxel segmentation method and device |
| CN108171131B (en) * | 2017-12-15 | 2022-01-14 | 湖北大学 | Improved MeanShift-based method for extracting Lidar point cloud data road marking line |
| CN108305279B (en) * | 2017-12-27 | 2019-07-05 | 东南大学 | A kind of super voxel generation method of the brain magnetic resonance image of iteration space fuzzy clustering |
| CN108364288B (en) * | 2018-03-01 | 2022-04-05 | 北京航空航天大学 | Segmentation method and device for breast cancer pathological image |
| CN108596209A (en) * | 2018-03-22 | 2018-09-28 | 重庆青信科技有限公司 | A kind of processing method and system of power spectrum CT images |
| CN108921816A (en) * | 2018-05-16 | 2018-11-30 | 东南大学 | A kind of cerebral function partition method based on super voxel and spectral clustering |
| US10964012B2 (en) | 2018-06-14 | 2021-03-30 | Sony Corporation | Automatic liver segmentation in CT |
| CN108961278B (en) * | 2018-06-20 | 2023-01-31 | 深圳市旭东数字医学影像技术有限公司 | Method and system for abdominal wall muscle segmentation based on image data |
| CN109191475B (en) * | 2018-09-07 | 2021-02-09 | 博志生物科技(深圳)有限公司 | Vertebral endplate segmentation method and device and computer readable storage medium |
| CN109461139B (en) * | 2018-09-29 | 2021-05-18 | 桂林电子科技大学 | Liver cancer quantitative analysis method based on dynamic MRI information fusion |
| CN111126242B (en) | 2018-10-16 | 2023-03-21 | 腾讯科技（深圳）有限公司 | Semantic segmentation method, device and equipment for lung image and storage medium |
| CN109872312B (en) * | 2019-02-15 | 2022-12-20 | 腾讯科技（深圳）有限公司 | Medical image segmentation method, device and system and image segmentation method |
| CN109949408A (en) * | 2019-03-13 | 2019-06-28 | 安徽紫薇帝星数字科技有限公司 | A kind of medical image method for reconstructing and its system cutting algorithm based on figure |
| CN113597631A (en) * | 2019-03-21 | 2021-11-02 | 柯惠有限合伙公司 | Organ segmentation method and system |
| CN109949322B (en) * | 2019-03-27 | 2023-06-23 | 中山大学 | Cerebrovascular image segmentation method based on magnetic resonance T1 enhanced image |
| CN110610497B (en) * | 2019-08-05 | 2023-06-16 | 史记生物技术(南京)有限公司 | Method for determining content of living pig carcass tissue based on CT image processing |
| CN111738991A (en) * | 2020-06-04 | 2020-10-02 | 西安数合信息科技有限公司 | Method for creating digital ray detection model of weld defects |
| CN113221691B (en) * | 2021-04-28 | 2022-09-27 | 展讯通信（上海）有限公司 | Fingerprint segmentation method and device, readable storage medium and terminal |
| CN114187320B (en) * | 2021-12-14 | 2022-11-08 | 北京柏惠维康科技股份有限公司 | Spine CT image segmentation method and spine imaging identification method and device |
| CN115067925B (en) * | 2022-06-14 | 2023-03-07 | 湖北省肿瘤医院（湖北省肿瘤研究所） | Liver magnetic resonance detection body position guiding method based on image processing |
| CN115359074B (en) * | 2022-10-20 | 2023-03-28 | 之江实验室 | Image segmentation and training method and device based on hyper-voxel clustering and prototype optimization |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102163327A (en) * | 2011-04-22 | 2011-08-24 | 陈宇珂 | Medical cardiac CT (computed tomography) image segmentation method |
| CN103824295A (en) * | 2014-03-03 | 2014-05-28 | 天津医科大学 | Segmentation method of adhesion vascular pulmonary nodules in lung CT (computed tomography) image |
| CN103942785A (en) * | 2014-04-09 | 2014-07-23 | 苏州大学 | PET and CT image lung tumor segmenting method based on graph cut |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080030497A1 (en) * | 2005-12-08 | 2008-02-07 | Yangqiu Hu | Three dimensional modeling of objects |
-
2015
- 2015-04-13 CN CN201510173307.6A patent/CN104809723B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102163327A (en) * | 2011-04-22 | 2011-08-24 | 陈宇珂 | Medical cardiac CT (computed tomography) image segmentation method |
| CN103824295A (en) * | 2014-03-03 | 2014-05-28 | 天津医科大学 | Segmentation method of adhesion vascular pulmonary nodules in lung CT (computed tomography) image |
| CN103942785A (en) * | 2014-04-09 | 2014-07-23 | 苏州大学 | PET and CT image lung tumor segmenting method based on graph cut |
Non-Patent Citations (2)
| Title |
|---|
| 医学CT图像分割方法研究;周生俊;《中国优秀博士学位论文全文数据库信息科技辑》;20131215;第32-33,37页 * |
| 基于SLIC与Delaunay图割的交互式图像分割算法;蔡强等;《计算机科学与探索》;20150410;第483-485页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104809723A (en) | 2015-07-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104809723B (en) | The three-dimensional CT image for liver automatic division method of algorithm is cut based on super voxel and figure | |
| Ponraj et al. | A survey on the preprocessing techniques of mammogram for the detection of breast cancer | |
| CN108364294B (en) | Multi-organ segmentation method for abdominal CT image based on superpixels | |
| CN109934235B (en) | Unsupervised abdominal CT sequence image multi-organ simultaneous automatic segmentation method | |
| CN106780518B (en) | A kind of MR image three-dimensional interactive segmentation method of the movable contour model cut based on random walk and figure | |
| CN105719295B (en) | A kind of intracranial hemorrhage region segmentation method and system based on three-dimensional super voxel | |
| CN107563378A (en) | The method and its system of area-of-interest are extracted in volume data | |
| CN109064476B (en) | CT chest radiography lung tissue image segmentation method based on level set | |
| CN104268873B (en) | Breast tumor partition method based on nuclear magnetic resonance images | |
| JP6539303B2 (en) | Transforming 3D objects to segment objects in 3D medical images | |
| CN107545579B (en) | Heart segmentation method, device and storage medium | |
| CN111667491B (en) | Breast tumor block diagram generation method with boundary marking information based on depth countermeasure network | |
| CN106780497B (en) | A kind of organ vascular tree extraction method based on statistical information | |
| CN109753997B (en) | Automatic accurate robust segmentation method for liver tumor in CT image | |
| CN105139377A (en) | Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver | |
| CN110866905A (en) | Rib identification and marking method | |
| CN105389821B (en) | It is a kind of that the medical image cutting method being combined is cut based on cloud model and figure | |
| CN106408576A (en) | Automatic region of interest segmentation method and system based on three-dimensional ultrasonic image | |
| CN104616289A (en) | Removal method and system for bone tissue in 3D CT (Three Dimensional Computed Tomography) image | |
| CN110458859B (en) | Multi-sequence MRI-based multiple myeloma focus segmentation system | |
| CN107169975B (en) | The analysis method and device of ultrasound image | |
| CN102663762B (en) | The dividing method of symmetrical organ in medical image | |
| Maitra et al. | Accurate breast contour detection algorithms in digital mammogram | |
| Häme | Liver tumor segmentation using implicit surface evolution | |
| Borzooie et al. | Modified density-based data clustering for interactive liver segmentation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180119 |