CN105701799B - Divide pulmonary vascular method and apparatus from lung's mask image - Google Patents
Divide pulmonary vascular method and apparatus from lung's mask image Download PDFInfo
- Publication number
- CN105701799B CN105701799B CN201511030492.XA CN201511030492A CN105701799B CN 105701799 B CN105701799 B CN 105701799B CN 201511030492 A CN201511030492 A CN 201511030492A CN 105701799 B CN105701799 B CN 105701799B
- Authority
- CN
- China
- Prior art keywords
- lung
- voxel
- pulmonary vascular
- image
- scale
- 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.)
- Active
Links
Classifications
-
- 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]
-
- 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
- G06T2207/20—Special algorithmic details
- G06T2207/20212—Image combination
- G06T2207/20221—Image fusion; Image merging
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30061—Lung
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The present invention proposes that one kind dividing pulmonary vascular method and apparatus from lung's mask image, this includes calculating the response of blood vessel similarity function of each voxel in blood vessel range scale under different blood vessel scale in lung mask image in multiple voxels from pulmonary vascular method is divided in lung's mask image, and the pulmonary vascular image using response as each voxel under different blood vessel scale；The target blood scale of each voxel in lung's mask image, and the corresponding pulmonary vascular image of target blood scale for obtaining each voxel are obtained, using the target pulmonary vascular image as each voxel；The target pulmonary vascular image of multiple voxels is merged, and obtains the target pulmonary vascular image after fusion.It can effectively be promoted through the invention and divide pulmonary vascular accuracy from lung's mask image, and promoted and divide pulmonary vascular effect from lung's mask image.
Description
Technical field
The present invention relates to segmentation Pulmonary Vascular technical field, more particularly to it is a kind of divide from lung's mask image it is pulmonary vascular
Method and apparatus.
Background technology
Accurate Pulmonary Vascular segmentation is lung's computer aided detection and the important step in diagnosis (CAD) system, lung bolt
Plug automatic identification with need first to extract Pulmonary Vascular tissue in detection to reduce the range of lesion detection, the early detection of lung cancer with examine
Disconnected middle removal Pulmonary Vascular interference utilizes the pulmonary vascular distribution guiding lobe of the lung to reduce the false positive of Lung neoplasm detection in surgical navigational
Segmentation.But complicated tree structure is presented in Pulmonary Vascular, intrapulmonary includes 23 grades of branches, and caliber is at 20 microns to 15 millimeters
Variation in range.In CT images, Pulmonary Vascular generally shows high density shadow because of the full blood in inside, but intensity profile is not
Uniformly, especially minute blood vessel is affected by partial volume effect.Tracheae, the Lung neoplasm and one of mucus are full of around Pulmonary Vascular
A little highly dense lesions can all interfere the accuracy that Pulmonary Vascular extracts.Pulmonary vascular geometrical model (Geometry Models) refers to blood
Manage the priori of the features of shape with elongated, tubulose, tree-shaped distribution.Hessian matrix characters analysis method can be effective
It identifies spherical object, cylindrical object, sheet object, is that one kind carries out the typical case of blood vessel segmentation using the geometrical model of blood vessel
Method.
The multiple dimensioned vessel filter device based on the analysis of Hessian matrix exgenvalues in the related technology, using all scales
Result of the maximum value of blood vessel similar function as scale selection.Under this mode, pulmonary vascular bifurcated junction response compared with
It is weak, it is discontinuous that blood vessel is be easy to cause after threshold division.
Invention content
The present invention is directed to solve at least some of the technical problems in related technologies.
For this purpose, an object of the present invention is to provide one kind dividing pulmonary vascular method, energy from lung's mask image
Enough effectively promoted divides pulmonary vascular accuracy from lung's mask image, and is promoted and divide lung blood from lung's mask image
The effect of pipe.
It is another object of the present invention to propose a kind of to divide pulmonary vascular device from lung's mask image.
In order to achieve the above objectives, what first aspect present invention embodiment proposed divides pulmonary vascular from lung's mask image
Method, including：Calculate the different blood vessel scale in blood vessel range scale of each voxel in lung's mask image in multiple voxels
Under blood vessel similarity function response, and the lung using the response as each voxel under different blood vessel scale
Portion's blood vessel image；The target blood scale of each voxel in lung's mask image is obtained, and obtains each voxel
The corresponding pulmonary vascular image of target blood scale, using the target pulmonary vascular image as each voxel；It will be described more
The target pulmonary vascular image of a voxel is merged, and obtains the target pulmonary vascular image after fusion.
What first aspect present invention embodiment proposed divides pulmonary vascular method from lung's mask image, by calculating lung
The response of blood vessel similarity function of each voxel in blood vessel range scale under different blood vessel scale in portion's mask image,
Using response as pulmonary vascular image, and the target pulmonary vascular image of each voxel in lung's mask image is obtained, and
The target pulmonary vascular image of multiple voxels is merged, can effectively be promoted divide from lung's mask image it is pulmonary vascular
Accuracy, and promoted and divide pulmonary vascular effect from lung's mask image.
In order to achieve the above objectives, what second aspect of the present invention embodiment proposed divides pulmonary vascular from lung's mask image
Device, including：Computing module, for calculating each voxel in lung's mask image in multiple voxels in blood vessel range scale
The response of blood vessel similarity function under different blood vessel scale, and using the response as each voxel in different blood
Pulmonary vascular image under pipe scale；Acquisition module, the target blood for obtaining each voxel in lung's mask image
Scale, and the corresponding pulmonary vascular image of target blood scale for obtaining each voxel, using as each voxel
Target pulmonary vascular image；Fusion Module for merging the target pulmonary vascular image of the multiple voxel, and obtains
Target pulmonary vascular image after fusion.
What second aspect of the present invention embodiment proposed divides pulmonary vascular device from lung's mask image, by calculating lung
The response of blood vessel similarity function of each voxel in blood vessel range scale under different blood vessel scale in portion's mask image,
Using response as pulmonary vascular image, and the target pulmonary vascular image of each voxel in lung's mask image is obtained, and
The target pulmonary vascular image of multiple voxels is merged, can effectively be promoted divide from lung's mask image it is pulmonary vascular
Accuracy, and promoted and divide pulmonary vascular effect from lung's mask image.
The additional aspect of the present invention and advantage will be set forth in part in the description, and will partly become from the following description
Obviously, or practice through the invention is recognized.
Description of the drawings
Above-mentioned and/or additional aspect and advantage of the invention will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that, wherein：
Fig. 1 is the flow signal for dividing pulmonary vascular method from lung's mask image that one embodiment of the invention proposes
Figure；
Fig. 2 is the flow signal for dividing pulmonary vascular method from lung's mask image that another embodiment of the present invention proposes
Figure；
Fig. 3 a are topical pulmonary mask image original image schematic diagrames；
Fig. 3 b are blood vessel scale σ filtered pulmonary vascular image schematic diagram when being 0.5mm；
Fig. 3 c are blood vessel scale σ filtered pulmonary vascular image schematic diagram when being 0.75mm；
Fig. 4 is that Pulmonary Vascular divides three-dimensional visualization design sketch in the embodiment of the present invention；
Fig. 5 is the structural representation for dividing pulmonary vascular device from lung's mask image that another embodiment of the present invention proposes
Figure；
Fig. 6 is the structural representation for dividing pulmonary vascular device from lung's mask image that another embodiment of the present invention proposes
Figure；
Fig. 7 is the structural representation for dividing pulmonary vascular device from lung's mask image that another embodiment of the present invention proposes
Figure；
Fig. 8 is the structural representation for dividing pulmonary vascular device from lung's mask image that another embodiment of the present invention proposes
Figure.
Specific implementation mode
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to attached
The embodiment of figure description is exemplary, and is only used for explaining the present invention, and is not considered as limiting the invention.On the contrary, this
The embodiment of invention includes all changes fallen within the scope of the spiritual and intension of attached claims, modification and is equal
Object.
Fig. 1 is the flow signal for dividing pulmonary vascular method from lung's mask image that one embodiment of the invention proposes
Figure, pulmonary vascular method should be divided from lung's mask image includes：
S101：Calculate the different blood vessel ruler in blood vessel range scale of each voxel in lung's mask image in multiple voxels
The response of blood vessel similarity function under degree, and the pulmonary vascular using response as each voxel under different blood vessel scale
Image.
Optionally, lung's mask image may be used the prior art to lung image carry out pretreatment get, herein not
It repeats again.
In an embodiment of the present invention, lung image can be the three-dimensional CT images of lung, by being carried out to lung image
Pretreatment obtains lung's mask image, and is split to Pulmonary Vascular according to multiple voxels in lung's mask image, can avoid
With the interference of the incoherent voxel of Pulmonary Vascular in lung image, the efficiency of Pulmonary Vascular segmentation is promoted.
Wherein, the three-dimensional CT images of lung are by CT scan (Computed Tomography, CT)
The image of the human lung scanned.
Voxel is the abbreviation of volume element, is least unit of the numerical data on three dimensions is divided, and voxel is used for three
The fields such as dimension imaging, science data and medical image.
In an embodiment of the present invention, voxel refers to pulmonary vascular volume element in lung's mask image.
In an embodiment of the present invention, blood vessel range scale is the position where each voxel in lung's mask image
The possible scale of Pulmonary Vascular, blood vessel range scale can be pre-assigned by the user, for example, blood vessel range scale is σ ∈
{σ1,...,σi,...,σk, i belongs to 1~k, and the value of k is 1~N, and N is positive integer.
It is alternatively possible to calculate blood vessel similarity function of each voxel in blood vessel range scale under each blood vessel scale
Response, and the pulmonary vascular image using response as each voxel under each blood vessel scale.
For example, blood vessel range scale is (A, B, C), then need to calculate blood vessel of the voxel 1 at A, tri- kinds of blood vessel scales of B, C
The response of similarity function, and calculate the response of blood vessel similarity function of the voxel 2 at A, tri- kinds of blood vessel scales of B, C
Value, and so on, calculate blood vessel similarity function of each voxel in lung's mask image at A, tri- kinds of blood vessel scales of B, C
Response.
Blood vessel similarity function is also referred to as vessel filter device, and vessel filter device is for example, Frangi filters, Huo Zhefen
Number rank differential filter, the embodiment of the present invention are not restricted this.
S102：The target blood scale of each voxel in lung's mask image is obtained, and obtains the target of each voxel
The corresponding pulmonary vascular image of blood vessel scale, using the target pulmonary vascular image as each voxel.
It is alternatively possible to optimize according to based on multiple labeling markov random file (Markov Random Field, MRF)
Method obtains the target blood scale of each voxel in lung's mask image.
In an embodiment of the present invention, the target blood scale of each voxel, i.e. blood vessel scale selection are obtained, it is each to be
Voxel is in blood vessel range scale σ ∈ { σ1,...,σi,...,σkIn selection one closest to the true ruler of Pulmonary Vascular where the voxel
The blood vessel scale of degree, as target blood scale, wherein i belongs to 1~k, and the value of k is 1~N, and N is positive integer.
Wherein, blood vessel scale choosing is carried out to each voxel in lung's mask image based on multiple labeling MRF optimization methods
It selects, is the partition problem that blood vessel scale selection problem is seen to mapping, blood vessel scale selection problem based on multiple labeling MRF optimization methods
Graph-theoretical Approach can be used (such as:Minimal cut) it solves.
In an embodiment of the present invention, it is a kind of energy minimization side based on graph theory based on multiple labeling MRF optimization methods
Method, its main thought are that the vertex that the voxel maps of lung's mask image are weighted graph will abut against the relationship between voxel
It is mapped as the side of weighted graph, and will abut against the weights that the similarity map between voxel is side, to obtain lung's mask image
Weighted graph, establish the energy model of weighted graph, cutting to figure completed by minimizing energy model.
In an embodiment of the present invention, target blood scale is a blood in the preassigned blood vessel range scale of user
Pipe scale, the true scale of Pulmonary Vascular of target blood scale closest to the current voxel position of each voxel, lung's mask
Each voxel in image corresponds to a target blood scale, and different voxels can correspond to difference in lung's mask image, or
The identical target blood scale of person.
In an embodiment of the present invention, the target pulmonary vascular image of each voxel corresponds to the target blood ruler of the voxel
Degree, the corresponding target pulmonary vascular image of target blood scale of each voxel are acquired by step S101.
S103：The target pulmonary vascular image of multiple voxels is merged, and obtains the target pulmonary vascular after fusion
Image.
It in an embodiment of the present invention, can be by the target lung blood of each voxel in multiple voxels in lung's mask image
Pipe image is merged, to obtain the target pulmonary vascular image after merging.
For example, including voxel 1, voxel 2 and voxel 3, and the target pulmonary vascular shadow of voxel 1 in lung's mask image
As being P1, the target pulmonary vascular image of voxel 2 is P2 and the target pulmonary vascular image of voxel 3 is P3, then by target lung
Portion blood vessel image P1, target pulmonary vascular image P2 and target pulmonary vascular image P3 are merged, after obtaining fusion
Target pulmonary vascular image P, and so on.
In the present embodiment, by calculating the different blood vessel ruler in blood vessel range scale of each voxel in lung's mask image
The response of blood vessel similarity function under degree using response as pulmonary vascular image, and obtains every in lung's mask image
The target pulmonary vascular image of a voxel, and the target pulmonary vascular image of multiple voxels is merged, it can effectively carry
It rises and divides pulmonary vascular accuracy from lung's mask image, and promoted and divide pulmonary vascular effect from lung's mask image
Fruit.
Fig. 2 is the flow signal for dividing pulmonary vascular method from lung's mask image that another embodiment of the present invention proposes
Figure, pulmonary vascular method should be divided from lung's mask image includes：
S201：Lung image is pre-processed, to obtain lung's mask image.
It is alternatively possible to be pre-processed to lung image using the prior art, to obtain lung's mask image.
In an embodiment of the present invention, lung image can be the three-dimensional CT images of lung, by being carried out to lung image
Pretreatment obtains lung's mask image, and is split to Pulmonary Vascular according to multiple voxels in lung's mask image, can avoid
With the interference of the incoherent voxel of Pulmonary Vascular in lung image, the efficiency of Pulmonary Vascular segmentation is promoted.
S202：Calculate the different blood vessel ruler in blood vessel range scale of each voxel in lung's mask image in multiple voxels
The response of blood vessel similarity function under degree, and the pulmonary vascular using response as each voxel under different blood vessel scale
Image.
It is alternatively possible to using the prior art to the gray level function of lung's mask image some voxel location parameter
p0The neighborhood Taylor expansion at place calculates each to analyze to obtain Hessian matrix exgenvalues according to blood vessel similarity function
The response of blood vessel similarity function of the voxel in blood vessel range scale under different blood vessel scale, and using response as each
Pulmonary vascular image of the voxel under different blood vessel scale, wherein blood vessel similarity function is also referred to as vessel filter device, blood
Pipe filter is not for example, Frangi filters or fractional order differential filter, the embodiment of the present invention are restricted this.
Fig. 3 is lung's blood vessel image filter effect figure under different blood vessel scale, wherein Fig. 3 a are topical pulmonary mask images
Original image schematic diagram, Fig. 3 b are blood vessel scale σ filtered pulmonary vascular image schematic diagram when being 0.5mm, Fig. 3 c are blood vessels
Scale σ filtered pulmonary vascular image schematic diagrames when being 0.75mm, from Fig. 3 b, Fig. 3 c it can be seen that when using compared with thin vessels
When scale, blood vessel similarity function can detect that thinner Pulmonary Vascular, and when using larger blood vessel scale, blood vessel similitude letter
Number can detect that thicker Pulmonary Vascular, in certain blood vessel range scale, Pulmonary Vascular of the blood vessel similarity function to different thicknesses
There is response.
S203：By the vertex that the voxel maps of lung's mask image are weighted graph, the relationship map between voxel will abut against
For the side of weighted graph, and will abut against the weights that the similarity map between voxel is side, with obtain lung's mask image plus
Weight graph.
In an embodiment of the present invention, lung mask image I (x, y, z) can be mapped as to weighted graph G (V, E), weighted
Vertex v ∈ V in figure correspond to the voxel in lung mask image I (x, y, z), the side in weighted graphMark is arbitrary
The syntople of two voxels (p, q), the weight on side identify the otherness between voxel, such as gray scale, position.
By the way that lung's mask image is mapped as weighted graph as, target blood scale selection problem is regarded to the division of weighted graph
Problem carries out space constraint to lung's mask image regional area attribute, can obtain the blood vessel size distribution of global optimization.
S204：The energy model of weighted graph is established, and obtains label when energy model being made to obtain energy-minimum.
In an embodiment of the present invention, the expression formula of energy model is as follows：
Wherein,For data item,For smooth item, η is voxel self-energy and adjacent when adjusting segmentation
Voxel connects the weight of energy, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, vector f corresponds to
The segmentation result of lung's mask image.
In an embodiment of the present invention, data item is set as：
Wherein, σiFor i-th of blood vessel scale in blood vessel range scale,It is p's for location parameter
In the response of the blood vessel similarity function of voxel, the response of maximum blood vessel similarity function, j=0 ..., k, the value of k
It is positive integer, L (p, σ for 1~N, Ni) be location parameter be p voxel in blood vessel scale σiUnder blood vessel similarity function sound
It should be worth；
Smooth item is set as：
Wherein, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, fqIt is covered to distribute to lung
Location parameter is the label of the voxel of q in film image.
In energy model, smooth itemThe interaction between adjacent voxels is defined, can be the function of labelIt can also be the function of basic dataAlternatively, can also be label and the function of basic dataSince adjacent voxels usually distribute similar or equal label, common smooth item is markd
Function is for example | fp-fq|, (fp-fq)2, min T, | fp-fq|, min { T, (fp-fq)2Etc., the function such as exp of basic data
(-(I(p)-I(q))2), min T, | fp-fq|, min { T, (fp-fq)2In, T is threshold value, is rule of thumb selected by user,
The embodiment of the present invention is not restricted this.
Specifically, similar according to blood vessel of the voxel that location parameter in lung's mask image is p under multiple blood vessel scales
Property function response in, the response and location parameter of maximum blood vessel similarity function are the voxel of p in blood vessel scale
σiUnder blood vessel similarity function response obtain with blood vessel scale σiCorresponding data item, and according to distributing to lung's mask
Location parameter is the label f of the voxel of p in imagepWith the label f for distributing to the voxel that location parameter is qqSmooth item is obtained, is led to
Cross calculating E (f), with obtain make energy model obtain energy-minimum when label, i be blood vessel scale index, i=0 ...,
The value of k, k are 1~N, and N is positive integer.
S205：Corresponding blood vessel scale will be marked as the target blood scale of each voxel in lung's mask image.
In an embodiment of the present invention, the target blood scale of each voxel, i.e. blood vessel scale selection are obtained, it is each to be
Voxel is in blood vessel range scale σ ∈ { σ1,...,σi,...,σkIn selection one closest to the true ruler of Pulmonary Vascular where the voxel
The blood vessel scale of degree, as target blood scale, wherein i belongs to 1~k, and the value of k is 1~N, and N is positive integer.
It is that energy-minimum is solved based on max-flow min-cut method based on multiple labeling MRF optimization methods, makes the cut set of figure
Cost be exactly equal to given energy model.Max-flow min-cut method includes mainly two major classes：Promote heavy label (Push
Relabel) method and augmenting path (Augmenting paths) method being changed using what Boykov was proposed in the embodiment of the present invention
Into augmenting path method the optimal solution of multiple labeling problem can be approached in polynomial time, the embodiment of the present invention does not make this
Limitation.
In an embodiment of the present invention, the energy model E (f) defined in step S204 solves energy model E (f) most
Small value is converted into the minimal cut set for seeking figure G (V, E).Segmentation result is exactly lung determined by the corresponding binary set of minimal cut set
Mask image segmentation result.
In an embodiment of the present invention, by obtaining each voxel in different blood vessels based on multiple labeling MRF optimization methods
The energy model of lung's mask image is set to obtain the blood vessel scale of energy-minimum in scale, to obtain the target lung of each voxel
Portion's blood vessel image is modeled for the relationship of neighbours' voxel, thus the blood vessel similitude letter with similar structure neighborhood voxel
Several responses and target blood scale may be not quite similar, and effectively promote the accuracy of Pulmonary Vascular segmentation.
S206：The corresponding pulmonary vascular image of target blood scale for obtaining each voxel, using the mesh as each voxel
Mark pulmonary vascular image.
In an embodiment of the present invention, it can be extracted in pulmonary vascular image of each voxel under different blood vessel scale
Go out pulmonary vascular image corresponding with target blood scale, obtains the target pulmonary vascular image of each voxel.
S207：The target pulmonary vascular image of multiple voxels is merged, and obtains the target pulmonary vascular after fusion
Image.
Specifically, the target pulmonary vascular image of each voxel extracted in step S206 is merged, to obtain
Target pulmonary vascular image after fusion.
S208：Judge the response of the blood vessel similarity function of each voxel in the target pulmonary vascular image after fusion
Whether predetermined threshold value is less than, if so, S209 is thened follow the steps, it is no to then follow the steps S210.
Be in this step by according to the response lag of blood vessel similarity function to the target pulmonary vascular image after fusion
The processing analyzed based on threshold value and degree of communication is carried out, for low density voxel and pulmonary vascular blood vessel similarity function response
When low, using the response lag of lower blood vessel similarity function, more ramuscule Pulmonary Vasculars can be partitioned into.
Wherein, predetermined threshold value is calculated by second function and is obtained, and second function is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, T (σp) be
The response lag of blood vessel similarity function, tσFor blood vessel scale threshold value, n is that blood vessel range scale medium vessels scale is less than tσBlood
The number of pipe scale, i are the index of blood vessel scale, and the value of i=0 ..., k, k are 1~N, and N is positive integer, TminAnd TmaxFor with
The specified tubulose characteristic response intensity threshold in family, TminSuch as it could be provided as 0.05, TmaxSuch as it could be provided as 0.15.
S209：The first function value of voxel is set as 0.
Wherein, first function is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, Vseg(L
(p,σp),σp) it is that Pulmonary Vascular divides discriminant function, L (p, σp) be location parameter be p voxel in target blood scale σpUnder
The response of blood vessel similarity function, T (σp) it is predetermined threshold value.
S210：The first function value of voxel is set as 1.
In an embodiment of the present invention, as shown in table 1 below, by Pulmonary Vascular segmentation result compared with goldstandard, the standard of segmentation
True property is described by three qualitative assessment indexs：ROC (Receiver Operating Characteristic) area under a curve
(AZ), specific (Specificity) and sensibility (Sensitivity).
Table 1
Pulmonary Vascular segmentation effect in the embodiment of the present invention is as shown in figure 4, Fig. 4 is Pulmonary Vascular segmentation three in the embodiment of the present invention
Effect of visualization figure is tieed up, Fig. 4 can be seen that the present invention can preferably be partitioned into the detail section in image, and be not susceptible to small
The interference of noise shows the less anatomical features of blood vessel in lung rift portion.
In the present embodiment, lung's mask image is obtained by being pre-processed to lung image, and according to lung's mask shadow
Multiple voxels are split Pulmonary Vascular as in, can avoid the interference with the incoherent voxel of Pulmonary Vascular in lung image, carry
Rise the efficiency of Pulmonary Vascular segmentation.Make lung in different blood vessel scale by being based on each voxel of multiple labeling MRF optimization methods acquisition
The energy model of portion's mask image obtains the blood vessel scale of energy-minimum, to obtain the target pulmonary vascular shadow of each voxel
Target blood scale selection problem is regarded as the partition problem of weighted graph by picture by the way that lung's mask image is mapped as weighted graph,
Space constraint is carried out to lung's mask image regional area attribute, the blood vessel size distribution of global optimization can be obtained.Pass through basis
What the response lag of blood vessel similarity function to the target pulmonary vascular image after fusion analyze based on threshold value and degree of communication
Processing, when for low density voxel and low pulmonary vascular blood vessel similarity function response, using lower blood vessel similitude
The response lag of function can be partitioned into more ramuscule Pulmonary Vasculars.Existed by calculating each voxel in lung's mask image
The response of blood vessel similarity function in blood vessel range scale under different blood vessel scale, using response as pulmonary vascular shadow
Picture, and obtain the target pulmonary vascular image of each voxel in lung's mask image, and by the target lung blood of multiple voxels
Pipe image is merged, and can effectively be promoted and be divided pulmonary vascular accuracy from lung's mask image, and is promoted from lung
Divide pulmonary vascular effect in mask image.
Fig. 5 is the structural representation for dividing pulmonary vascular device from lung's mask image that another embodiment of the present invention proposes
Figure, it includes computing module 501 that pulmonary vascular device 50 should be divided from lung's mask image, for calculating in lung's mask image
The response of blood vessel similarity function of each voxel in blood vessel range scale under different blood vessel scale in multiple voxels, and
Pulmonary vascular image using response as each voxel under different blood vessel scale；Acquisition module 502 is covered for obtaining lung
The target blood scale of each voxel in film image, and the corresponding pulmonary vascular shadow of target blood scale for obtaining each voxel
Picture, using the target pulmonary vascular image as each voxel；Fusion Module 503 is used for the target pulmonary vascular of multiple voxels
Image is merged, and obtains the target pulmonary vascular image after fusion.
Computing module 501, for calculating each voxel in lung's mask image in multiple voxels in blood vessel range scale
The response of blood vessel similarity function under middle different blood vessel scale, and using response as each voxel in different blood vessel scale
Under pulmonary vascular image.
Acquisition module 502, the target blood scale for obtaining each voxel in lung's mask image, and obtain per individual
The corresponding pulmonary vascular image of target blood scale of element, using the target pulmonary vascular image as each voxel.
Optionally, as shown in fig. 6, acquisition module 502 specifically includes：
Target blood scale acquisition submodule 5021, for being obtained according to based on multiple labeling markov random file optimization method
Take the target blood scale of each voxel in lung's mask image.
Optionally, as shown in fig. 7, target blood scale acquisition submodule 5021 includes：
Weighted graph map unit 50211 is covered for being weighted figure mapping processing to lung's mask image with obtaining lung
The weighted graph of film image.
Optionally, weighted graph map unit 50211 is specifically used for：
By the vertex that the voxel maps of lung's mask image are weighted graph, will abut against the relationship map between voxel is weighting
The side of figure, and the weights that the similarity map between voxel is side are will abut against, to obtain the weighted graph of lung's mask image.
Energy model establishes unit 50212, the energy model for establishing weighted graph, and obtaining makes energy model obtain energy
Measure label when minimum value.
Optionally, the energy model of weighted graph is：
Wherein,For data item,For smooth item, η is voxel self-energy and adjacent when adjusting segmentation
Voxel connects the weight of energy, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, vector f corresponds to
The segmentation result of lung's mask image.
Optionally, data item is set as：
Wherein, σiFor i-th of blood vessel scale in blood vessel range scale,It is p's for location parameter
In the response of the blood vessel similarity function of voxel, the response of maximum blood vessel similarity function, j=0 ..., k, the value of k
It is positive integer, L (p, σ for 1~N, Ni) be location parameter be p voxel in blood vessel scale σiUnder blood vessel similarity function sound
It should be worth；
Smooth item is set as：
Wherein, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, fqIt is covered to distribute to lung
Location parameter is the label of the voxel of q in film image.
Acquiring unit 50213, for corresponding blood vessel scale will to be marked as the mesh of each voxel in lung's mask image
Mark blood vessel scale.
Target pulmonary vascular image capturing submodule 5022, the corresponding lung of target blood scale for obtaining each voxel
Portion's blood vessel image, using the target pulmonary vascular image as each voxel.
Fusion Module 503 for merging the target pulmonary vascular image of multiple voxels, and obtains the mesh after fusion
Mark pulmonary vascular image.
Optionally, as shown in figure 8, this be divided pulmonary vascular device 50 from lung's mask image and further include：
Comparison module 504, the blood vessel similitude letter for each voxel in the target pulmonary vascular image after merging
Several responses are compared with predetermined threshold value.
Optionally, predetermined threshold value is calculated by second function and is obtained, and second function is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, T (σp) be
The response lag of blood vessel similarity function, tσFor blood vessel scale threshold value, n is that blood vessel range scale medium vessels scale is less than tσBlood
The number of pipe scale, i are the index of blood vessel scale, and the value of i=0 ..., k, k are 1~N, and N is positive integer, TminAnd TmaxFor with
The specified tubulose characteristic response intensity threshold in family.
Setup module 505, the response of the blood vessel similarity function for each voxel in target pulmonary vascular image
When value is less than predetermined threshold value, the first function value of voxel is set as 0, the blood of each voxel in target pulmonary vascular image
When the response of pipe similarity function is greater than or equal to predetermined threshold value, the first function value of voxel is set as 1.
Optionally, first function is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, Vseg(L
(p,σp),σp) it is that Pulmonary Vascular divides discriminant function, L (p, σp) be location parameter be p voxel in target blood scale σpUnder
The response of blood vessel similarity function, T (σp) it is predetermined threshold value.
Optionally, as shown in figure 8, this be divided pulmonary vascular device 50 from lung's mask image and further include：
Preprocessing module 506, for being pre-processed to lung image, to obtain lung's mask image.
It should be noted that the aforementioned explanation to pulmonary vascular dividing method embodiment is also applied for the embodiment
Pulmonary vascular segmenting device 50, realization principle is similar, and details are not described herein again.
In the present embodiment, by calculating the different blood vessel ruler in blood vessel range scale of each voxel in lung's mask image
The response of blood vessel similarity function under degree using response as pulmonary vascular image, and obtains every in lung's mask image
The target pulmonary vascular image of a voxel, and the target pulmonary vascular image of multiple voxels is merged, it can effectively carry
It rises and divides pulmonary vascular accuracy from lung's mask image, and promoted and divide pulmonary vascular effect from lung's mask image
Fruit.
It should be noted that in the description of the present invention, term " first ", " second " etc. are used for description purposes only, without
It can be interpreted as indicating or implying relative importance.In addition, in the description of the present invention, unless otherwise indicated, the meaning of " multiple "
It is two or more.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes
It is one or more for realizing specific logical function or process the step of executable instruction code module, segment or portion
Point, and the range of the preferred embodiment of the present invention includes other realization, wherein can not press shown or discuss suitable
Sequence, include according to involved function by it is basic simultaneously in the way of or in the opposite order, to execute function, this should be of the invention
Embodiment person of ordinary skill in the field understood.
It should be appreciated that each section of the present invention can be realized with hardware, software, firmware or combination thereof.Above-mentioned
In embodiment, software that multiple steps or method can in memory and by suitable instruction execution system be executed with storage
Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware
Any one of row technology or their combination are realized：With the logic gates for realizing logic function to data-signal
Discrete logic, with suitable combinational logic gate circuit application-specific integrated circuit, programmable gate array (PGA), scene
Programmable gate array (FPGA) etc..
Those skilled in the art are appreciated that realize all or part of step that above-described embodiment method carries
Suddenly it is that relevant hardware can be instructed to complete by program, the program can be stored in a kind of computer-readable storage medium
In matter, which includes the steps that one or a combination set of embodiment of the method when being executed.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, it can also
That each unit physically exists alone, can also two or more units be integrated in a module.Above-mentioned integrated mould
The form that hardware had both may be used in block is realized, can also be realized in the form of software function module.The integrated module is such as
Fruit is realized in the form of software function module and when sold or used as an independent product, can also be stored in a computer
In read/write memory medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show
The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example
Point is included at least one embodiment or example of the invention.In the present specification, schematic expression of the above terms are not
Centainly refer to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be any
One or more embodiments or example in can be combined in any suitable manner.
Although the embodiments of the present invention has been shown and described above, it is to be understood that above-described embodiment is example
Property, it is not considered as limiting the invention, those skilled in the art within the scope of the invention can be to above-mentioned
Embodiment is changed, changes, replacing and modification.
Claims (16)
1. one kind dividing pulmonary vascular method from lung's mask image, which is characterized in that include the following steps：
Calculate blood of each voxel in lung's mask image in multiple voxels in blood vessel range scale under different blood vessel scale
The response of pipe similarity function, and the pulmonary vascular using the response as each voxel under different blood vessel scale
Image；
The target blood scale of each voxel in lung's mask image is obtained, and obtains the target blood of each voxel
The corresponding pulmonary vascular image of scale, using the target pulmonary vascular image as each voxel；
The target pulmonary vascular image of the multiple voxel is merged, and obtains the target pulmonary vascular image after fusion.
2. dividing pulmonary vascular method from lung's mask image as described in claim 1, which is characterized in that wherein, according to
The target blood scale of each voxel in lung's mask image is obtained based on multiple labeling markov random file optimization method.
3. dividing pulmonary vascular method from lung's mask image as claimed in claim 2, which is characterized in that described according to base
The target blood scale of each voxel in lung's mask image is obtained in multiple labeling markov random file optimization method, is wrapped
It includes：
Figure mapping processing is weighted to lung's mask image, to obtain the weighted graph of lung's mask image；
The energy model of the weighted graph is established, and obtains the label when energy model being made to obtain energy-minimum；
Using the corresponding blood vessel scale of the label as the target blood scale of each voxel in lung's mask image.
4. dividing pulmonary vascular method from lung's mask image as claimed in claim 3, which is characterized in that described to described
Lung's mask image is weighted figure mapping processing, to obtain the weighted graph of lung's mask image, including：
By the vertex that the voxel maps of lung's mask image are the weighted graph, the relationship map that will abut against between voxel is
The side of the weighted graph, and by the weights that the similarity map between the adjacent voxel is the side, to obtain the lung
The weighted graph of portion's mask image.
5. dividing pulmonary vascular method from lung's mask image as claimed in claim 3, which is characterized in that the weighted graph
Energy model be：
Wherein,For data item,For smooth item, η is voxel self-energy and adjacent voxels when adjusting segmentation
Connect the weight of energy, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, vector f corresponds to
The segmentation result of lung's mask image.
6. dividing pulmonary vascular method from lung's mask image as claimed in claim 5, which is characterized in that wherein, described
Data item is set as：
Wherein, σiFor i-th of blood vessel scale in blood vessel range scale,The voxel for being p for location parameter
Blood vessel similarity function response in, the response of maximum blood vessel similarity function, j=0 ..., k, the value of k is 1
~N, N are positive integer, L (p, σi) be location parameter be p voxel in blood vessel scale σiUnder blood vessel similarity function response
Value；
The smooth item is set as：
Wherein, fpTo distribute to the label for the voxel that location parameter is p in lung's mask image, fqTo distribute to the lung
Location parameter is the label of the voxel of q in portion's mask image.
7. dividing pulmonary vascular method from lung's mask image as described in claim 1, which is characterized in that merged obtaining
After target pulmonary vascular image afterwards, further include：
By the response of the blood vessel similarity function of each voxel in the target pulmonary vascular image after the fusion with
Predetermined threshold value is compared；
If it is less than the predetermined threshold value, then the first function value of the voxel is set as 0；
If it is greater than or equal to the predetermined threshold value, then the first function value of the voxel is set as 1.
8. dividing pulmonary vascular method from lung's mask image as claimed in claim 7, which is characterized in that first letter
Number is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, Vseg(L
(p,σp),σp) it is that Pulmonary Vascular divides discriminant function, L (p, σp) be location parameter be p voxel in target blood scale σpUnder
The response of blood vessel similarity function, T (σp) it is the predetermined threshold value.
9. dividing pulmonary vascular method from lung's mask image as claimed in claim 7, which is characterized in that the default threshold
Value is calculated by second function and is obtained, and the second function is：
Wherein, p is the location parameter of each voxel, σpFor the target blood scale for the voxel that location parameter is p, T (σp) be
The response lag of the blood vessel similarity function, tσFor blood vessel scale threshold value, n is that blood vessel range scale medium vessels scale is less than tσ
Blood vessel scale number, i is the index of blood vessel scale, and i=0 ..., k, the value of k is 1~N, and N is positive integer, TminAnd Tmax
The tubulose characteristic response intensity threshold specified for user.
10. dividing pulmonary vascular method from lung's mask image as described in claim 1, which is characterized in that in the meter
Calculate blood vessel similitude letter of each of the multiple voxels in blood vessel range scale under different blood vessel scale in lung's mask image
Before several responses, further include：
Lung image is pre-processed, to obtain lung's mask image.
11. one kind dividing pulmonary vascular device from lung's mask image, which is characterized in that including：
Computing module, for calculating the different blood in blood vessel range scale of each voxel in lung's mask image in multiple voxels
The response of blood vessel similarity function under pipe scale, and using the response as each voxel in different blood vessel scale
Under pulmonary vascular image；
Acquisition module, the target blood scale for obtaining each voxel in lung's mask image, and obtain described each
The corresponding pulmonary vascular image of target blood scale of voxel, using the target pulmonary vascular image as each voxel；
Fusion Module for merging the target pulmonary vascular image of the multiple voxel, and obtains the target after fusion
Pulmonary vascular image.
12. dividing pulmonary vascular device from lung's mask image as claimed in claim 11, which is characterized in that the acquisition
Module specifically includes：
Target blood scale acquisition submodule, for obtaining the lung according to based on multiple labeling markov random file optimization method
The target blood scale of each voxel in portion's mask image；
Target pulmonary vascular image capturing submodule, the corresponding lung's blood of target blood scale for obtaining each voxel
Pipe image, using the target pulmonary vascular image as each voxel.
13. dividing pulmonary vascular device from lung's mask image as claimed in claim 12, which is characterized in that the target
Blood vessel scale acquisition submodule includes：
Weighted graph map unit is covered for being weighted figure mapping processing to lung's mask image with obtaining the lung
The weighted graph of film image；
Energy model establishes unit, the energy model for establishing the weighted graph, and obtaining makes the energy model obtain energy
Measure label when minimum value；
Acquiring unit, for using the corresponding blood vessel scale of the label as the target of each voxel in lung's mask image
Blood vessel scale.
14. dividing pulmonary vascular device from lung's mask image as claimed in claim 13, which is characterized in that the weighting
Figure map unit is specifically used for：
By the vertex that the voxel maps of lung's mask image are the weighted graph, the relationship map that will abut against between voxel is
The side of the weighted graph, and by the weights that the similarity map between the adjacent voxel is the side, to obtain the lung
The weighted graph of portion's mask image.
15. dividing pulmonary vascular device from lung's mask image as claimed in claim 11, which is characterized in that further include：
Comparison module is used for the blood vessel similitude letter of each voxel in the target pulmonary vascular image after the fusion
Several responses are compared with predetermined threshold value；
Setup module, the response of the blood vessel similarity function for each voxel in the target pulmonary vascular image
When value is less than the predetermined threshold value, the first function value of the voxel is set as 0, in the target pulmonary vascular image
When the response of the blood vessel similarity function of each voxel is greater than or equal to the predetermined threshold value, by the first of the voxel
Functional value is set as 1.
16. dividing pulmonary vascular device from lung's mask image as claimed in claim 11, which is characterized in that further include：
Preprocessing module, for being pre-processed to lung image, to obtain lung's mask image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511030492.XA CN105701799B (en) | 2015-12-31 | 2015-12-31 | Divide pulmonary vascular method and apparatus from lung's mask image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201511030492.XA CN105701799B (en) | 2015-12-31 | 2015-12-31 | Divide pulmonary vascular method and apparatus from lung's mask image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105701799A CN105701799A (en) | 2016-06-22 |
| CN105701799B true CN105701799B (en) | 2018-10-30 |
Family
ID=56226060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201511030492.XA Active CN105701799B (en) | 2015-12-31 | 2015-12-31 | Divide pulmonary vascular method and apparatus from lung's mask image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105701799B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018119766A1 (en) * | 2016-12-28 | 2018-07-05 | 上海联影医疗科技有限公司 | Multi-modal image processing system and method |
| CN106875398B (en) * | 2017-01-04 | 2020-06-19 | 北京数科网维技术有限责任公司 | Method, device and terminal for realizing interactive image segmentation |
| CN107633514B (en) * | 2017-09-19 | 2020-07-31 | 北京大学第三医院 | Pulmonary nodule peripheral blood vessel quantitative evaluation system and method |
| CN108010041B (en) * | 2017-12-22 | 2020-08-11 | 数坤（北京）网络科技有限公司 | Human heart coronary artery extraction method |
| CN108364297B (en) * | 2018-03-19 | 2020-08-14 | 青岛海信医疗设备股份有限公司 | Blood vessel image segmentation method, terminal and storage medium |
| CN109872328B (en) | 2019-01-25 | 2021-05-07 | 腾讯科技（深圳）有限公司 | Brain image segmentation method, device and storage medium |
| CN111383191B (en) * | 2019-12-11 | 2024-03-08 | 北京深睿博联科技有限责任公司 | Image processing method and device for vascular fracture repair |
| CN112465749B (en) * | 2020-11-11 | 2024-03-22 | 沈阳东软智能医疗科技研究院有限公司 | Method and device for extracting pulmonary embolism image, storage medium and electronic equipment |
| CN113361509B (en) * | 2021-08-11 | 2021-11-09 | 西安交通大学医学院第一附属医院 | Image processing method for facial paralysis detection |
| CN114202469B (en) * | 2021-11-11 | 2022-08-19 | 北京医准智能科技有限公司 | Method and device for selecting hyper-parameters of Frangi filter, electronic equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393644A (en) * | 2008-08-15 | 2009-03-25 | 华中科技大学 | Hepatic portal vein tree modeling method and system thereof |
| CN101425186A (en) * | 2008-11-17 | 2009-05-06 | 华中科技大学 | Liver subsection method based on CT image and system thereof |
| CN102243759A (en) * | 2010-05-10 | 2011-11-16 | 东北大学 | Three-dimensional lung vessel image segmentation method based on geometric deformation model |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| ATE314703T1 (en) * | 2001-10-16 | 2006-01-15 | Koninkl Philips Electronics Nv | METHOD FOR AUTOMATIC LABELING OF BRANCHES |
| US7058210B2 (en) * | 2001-11-20 | 2006-06-06 | General Electric Company | Method and system for lung disease detection |
| US7020316B2 (en) * | 2001-12-05 | 2006-03-28 | Siemens Corporate Research, Inc. | Vessel-feeding pulmonary nodule detection by volume projection analysis |
-
2015
- 2015-12-31 CN CN201511030492.XA patent/CN105701799B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101393644A (en) * | 2008-08-15 | 2009-03-25 | 华中科技大学 | Hepatic portal vein tree modeling method and system thereof |
| CN101425186A (en) * | 2008-11-17 | 2009-05-06 | 华中科技大学 | Liver subsection method based on CT image and system thereof |
| CN102243759A (en) * | 2010-05-10 | 2011-11-16 | 东北大学 | Three-dimensional lung vessel image segmentation method based on geometric deformation model |
Non-Patent Citations (5)
| Title |
|---|
| 《3D INTERACTIVE CORONARY ARTERY SEGMENTATION USING RANDOM FORESTS AND MARKOV RANDOM FIELD OPTIMIZATION》;Jingjing Deng等;《2014 IEEE International Conference on Image Processing》;20141231;全文 * |
| 《A Novel Approach for Lung Nodules Segmentation in Chest CT Using Level Sets》;Amal A.Farag等;《IEEE TRANSACTIONS ON IMAGE PROCESSING》;20131231;第22卷(第12期);全文 * |
| 《一种基于Canny算子的level-set肺部血管分割算法》;高齐新等;《***仿真学报》;20081031;第20卷(第20期);全文 * |
| 《一种基于水平集的肺部血管快速分割方法》;高齐新等;《东北大学学报（自然科学版）》;20080630;第29卷(第6期);全文 * |
| 《基于Markov随机场的脑部三维磁共振血管造影数据的分割》;周寿军等;《集成技术》;20140131;第3卷(第1期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105701799A (en) | 2016-06-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105701799B (en) | Divide pulmonary vascular method and apparatus from lung's mask image | |
| KR102204437B1 (en) | Apparatus and method for computer aided diagnosis | |
| EP3035287B1 (en) | Image processing apparatus, and image processing method | |
| US10111713B2 (en) | Surgery assistance apparatus, surgery assistance method and non-transitory computer-readable recording medium having stored therein surgery assistance program | |
| Maitra et al. | Technique for preprocessing of digital mammogram | |
| Camilus et al. | Computer-aided identification of the pectoral muscle in digitized mammograms | |
| JP4950071B2 (en) | Method for automatic extraction of pulmonary artery tree from 3D medical images | |
| US8605978B2 (en) | Medical image processing apparatus and method, and computer readable recording medium on which is recorded program for the same | |
| CN101027692B (en) | System and method for object characterization of toboggan-based clusters | |
| CN108133476B (en) | Method and system for automatically detecting pulmonary nodules | |
| Chang et al. | Graph-based learning for segmentation of 3D ultrasound images | |
| Mouton et al. | Materials-based 3D segmentation of unknown objects from dual-energy computed tomography imagery in baggage security screening | |
| KR20160020917A (en) | Apparatus and Method for computer aided diagnosis | |
| Huang et al. | Nucleus segmentation of cervical cytology images based on multi-scale fuzzy clustering algorithm | |
| KR20140009407A (en) | Method and apparatus for isolating a potential anomaly in imaging data and its application to medical imagery | |
| CN111899850A (en) | Medical image information processing method, display method and readable storage medium | |
| Rebouças Filho et al. | 3D segmentation and visualization of lung and its structures using CT images of the thorax | |
| Fiori et al. | A complete system for candidate polyps detection in virtual colonoscopy | |
| CN112541909B (en) | Lung nodule detection method and system based on three-dimensional neural network of slice perception | |
| CN114533002A (en) | Carotid artery central line extraction method and device, storage medium and electronic equipment | |
| Mourya et al. | Empirical greedy machine‐based automatic liver segmentation in CT images | |
| KR102227921B1 (en) | Qualification method of lung vessel based on lobe | |
| KR102187712B1 (en) | A Method For Merging Veins In A X-ray/CT Image And An Apparatus Thereof | |
| JP2017189384A (en) | Image processor, image processing method and program | |
| CN114631849A (en) | Abdominal aorta imaging method and related apparatus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |