CN108269261A - A kind of Bones and joints CT image partition methods and system - Google Patents
A kind of Bones and joints CT image partition methods and system Download PDFInfo
- Publication number
- CN108269261A CN108269261A CN201611263197.3A CN201611263197A CN108269261A CN 108269261 A CN108269261 A CN 108269261A CN 201611263197 A CN201611263197 A CN 201611263197A CN 108269261 A CN108269261 A CN 108269261A
- Authority
- CN
- China
- Prior art keywords
- class
- bony areas
- ellipsoid
- images
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
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
- G06T2207/20081—Training; Learning
-
- 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
- G06T2207/30008—Bone
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The application provides a kind of Bones and joints CT image partition methods and system, by being pre-processed to the CT images got, obtains the bony areas in the CT images;Class elliptic region positioning is carried out to obtained each bony areas, obtains being imaged as the bony areas of class ellipsoid;According to preset rules by selecting at least one class ellipsoid in obtained each class ellipsoid, the class ellipsoid of selection is labeled as seed region;It is background area by non-bone zone marker by the seed region labeled as target area.Using the energy smallest partition algorithm based on graph theory, by the target area and background area of above-mentioned setting, the bony areas for being connected to one piece in joint is separated, and different marks is set to different bony areas.The automatic segmentation of CT images is realized, improves the efficiency of CT images segmentation.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of Bones and joints CT image partition methods and system.
Background technology
It is largely needed both at home and abroad for Bones and joints CT image three-dimensionals volume segmentation method in each CT sequence chart at present
It is manuallyd locate or manual segmentation as in, it is time-consuming and laborious, and efficiency is low.Therefore be badly in need of one kind be used to implement to arthrosis image into
The dividing method of row Fast Segmentation.
Invention content
In view of this, the embodiment of the present invention provides a kind of Bones and joints CT image partition methods and system, to realize Bones and joints
The automatic segmentation of CT images.
To achieve the above object, the embodiment of the present invention provides following technical solution:
A kind of Bones and joints CT image partition methods, including:
To the CT images got, the CT images are pre-processed, obtain the bony areas in the CT images;
Class elliptic region positioning is carried out to obtained each bony areas using Hough transform, obtains being imaged as class
The bony areas of ellipsoid;
At least one class ellipsoid is selected from obtained each class ellipsoid according to preset rules, the class of selection is oval
Face is labeled as seed region;
It is background area by non-bone zone marker by the seed region labeled as target area;
It, will even by the target area and background area of above-mentioned label using the energy smallest partition algorithm based on graph theory
It is connected on one piece of bony areas to separate in joint, and different marks is set to different bony areas.
Preferably, in above-mentioned Bones and joints CT image partition methods, the pretreatment includes:Including:
The CT images got are positioned, obtain the human region in image;
Noise reduction process is carried out to the human region that positioning obtains;
Bony areas positioning is carried out to the human region after noise reduction process and removes soft tissue area, obtains the CT images
In bony areas.
Preferably, in above-mentioned Bones and joints CT image partition methods, the bony areas for being connected to one piece is separated in joint,
And it after different marks is set to different bony areas, further includes:
The boundary of different bony areas is optimized.
Preferably, it is described oval by obtained each class according to preset rules in above-mentioned Bones and joints CT image partition methods
At least one class ellipsoid is selected in face, including:
According to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6The score value S of each class ellipsoid is calculated, will divide
The highest one or more class ellipsoids of value S are labeled as seed region;
Wherein, the x1For the edge gray scale of class ellipsoid and center gray value ratio, the x2Major and minor axis for class ellipsoid
Than x3For the area of class ellipsoid, x4For the long axis of class ellipsoid and the angle of horizontal plane, x5For class ellipsoid central point with
The distance at target area center, the target area are according to the bony areas in binary-state threshold method labeling CT image, x6For
The boundary of class ellipsoid and the distance of soft tissue boundary line, the soft tissue boundary line are gray scale in the image greyscale histogram
It is worth the boundary formed in the bony areas for the position of preset value;The a1、a2、a3、a4、a5、a6Respectively predetermined coefficient.
Preferably, it in above-mentioned Bones and joints CT image partition methods, further includes:Obtain the highest classes of score value S input by user
The true score value S ' of ellipsoid, according to the true score value S ', x1、x2、x3、x4、x5And x6To formula S=a1x1+a2x2+a3x3+
a4x4+a5x5+a6x6It is trained, a is adjusted according to training result1、a2、a3、a4、a5And a6Size.
A kind of Bones and joints CT image segmentation systems, including:
Pretreatment unit, for the CT images got, being pre-processed to the CT images, obtaining the CT images
In bony areas;
Class elliptic region positioning unit, it is ellipse for carrying out class to obtained each bony areas using Hough transform
Circle zone location obtains being imaged as the bony areas of class ellipsoid;
Seed region positioning unit, for selecting at least one class from obtained each class ellipsoid according to preset rules
The class ellipsoid of selection is labeled as seed region by ellipsoid;
Indexing unit, for labeled as target area, the bony areas corresponding to the seed region will to be removed the bone
Zone marker other than bone region is background area;
Indexing unit, for using the energy smallest partition algorithm based on graph theory is utilized, passing through the target area of above-mentioned label
Domain and background area separate the bony areas for being connected to one piece in joint, and different to different bony areas settings
Mark.
Preferably, in above-mentioned Bones and joints CT image segmentation systems, the pretreatment unit is specifically configured to:
Position the human region in the CT images got;Noise reduction process is carried out to the human region that positioning obtains;To drop
Making an uproar treated, human region carries out bony areas positioning and removes soft tissue area, obtains the bone area in the CT images
Domain.
Preferably, it in above-mentioned Bones and joints CT image segmentation systems, further includes:
Optimize unit, for being optimized to the bony areas for setting different marks into row bound.
Preferably, in above-mentioned Bones and joints CT image segmentation systems, the seed region positioning unit is specifically configured to:
According to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6The score value S of each class ellipsoid is calculated, will divide
The highest one or more class ellipsoids of value S are labeled as seed region;
Wherein, the x1For the edge gray scale of class ellipsoid and center gray value ratio, the x2Major and minor axis for class ellipsoid
Than x3For the area of class ellipsoid, x4For the long axis of class ellipsoid and the angle of horizontal plane, x5For class ellipsoid central point with
The distance of target area central point, the target area are according to the bony areas in threshold method labeling CT image, x6It is ellipse for class
The boundary of disc and the distance of soft tissue boundary line, the soft tissue boundary line is that gray value is in the image greyscale histogram
The boundary that the position of preset value is formed in the bony areas;The a1、a2、a3、a4、a5、a6Respectively predetermined coefficient.
Preferably, it in above-mentioned Bones and joints CT image segmentation systems, further includes:
Formula optimization unit, for obtaining the true score value S ' of the highest class ellipsoids of score value S input by user, according to institute
State true score value S ', the x of the highest class ellipsoids of score value S1、x2、x3、x4、x5And x6To formula S=a1x1+a2x2+a3x3+a4x4+
a5x5+a6x6It is trained, a is adjusted according to training result1、a2、a3、a4、a5And a6Size.
The application provides a kind of Bones and joints CT image partition methods and system, by being located in advance to the CT images got
Reason, obtains the bony areas in the CT images;Class elliptic region positioning is carried out to obtained each bony areas, is obtained
It is imaged as the bony areas of class ellipsoid;Select at least one class oval from obtained each class ellipsoid according to preset rules
The class ellipsoid of selection is labeled as seed region by face;By the seed region labeled as target area, by non-bony areas mark
It is denoted as background area.Using the partitioning algorithm based on graph theory, by the target area and background area of above-mentioned setting, will be connected to
One piece of bony areas is separated in joint, and different marks is set to different bony areas.Realize oneself of CT images
Dynamic segmentation improves the efficiency of CT images segmentation.
Automatic precision refinement segmentation is realized using the energy smallest partition algorithm based on graph theory, take full advantage of in three dimensions
The algorithmic characteristic using energy minimization path as boundary, can reach satisfied segmentation effect to a certain extent.
And in system in actual use, can by inputting training sample, based on self-learning algorithm adjustment formula S=
a1x1+a2x2+a3x3+a4x4+a5x5+a6x6In predetermined coefficient a1、a2、a3、a4、a5、a6So that the accuracy of seed region positioning
It is continuously improved, so as to improve the precision of segmentation.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
The embodiment of invention, for those of ordinary skill in the art, without creative efforts, can also basis
The attached drawing of offer obtains other attached drawings.
Fig. 1 is a kind of flow diagram of Bones and joints CT image partition methods disclosed in the embodiment of the present application;
Fig. 2 is the internal anatomy of hip joint;
Fig. 3 is basin bone CT images;
Fig. 4 is label human region image;
Fig. 5 is label bone region image;
Fig. 6 is the image of class ellipse fitting positioning seed region;
Fig. 7 is GraphCut algorithm flow charts;
Fig. 8 is last segmentation result label figure;
Fig. 9 is segmentation result three-dimensional reconstruction design sketch;
Figure 10 is a kind of image greyscale histogram schematic diagram of bony areas;
Figure 11 is a kind of structure diagram of Bones and joints CT image segmentation systems disclosed in the embodiment of the present application.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without making creative work
Embodiment shall fall within the protection scope of the present invention.
CT is that the certain thickness level in human body portion is scanned with X-ray beam, is received by detector through the level
X ray, after being changed into visible ray, electric signal is become, then through analog/digital converter (analog/ from opto-electronic conversion
Digital converter) switch to digital signal, input computer disposal.The processing that image is formed is like to selected level point
The cuboid identical into several volumes, referred to as voxel (voxel).Different tissues also has in CT imagings upper density and compares
Big difference, atmospheric density is than relatively low under normal circumstances, and secondly, blood vessel is higher, bone or calcified tissue's density highest for fat.
CT images joint segmentation refer to come out joint separate marking in CT images, be mainly used for it is preoperative planning or
Person's lesion analysis etc., such as hip joint refer to coming out hip joint separate marking in CT images.
Referring to Fig. 1, this application discloses a kind of Bones and joints CT image partition methods, to realize to Bones and joints in CT images
Automatic segmentation, such as the segmentation in the CT images of hipbone joint to hip joint, referring to Fig. 1, this method includes:
Step S101:Obtain CT images;
In this step, the CT images got can be the scan image of a sequence as shown in Figure 3, each sequence
Can include 200-300 CT images;
Step S102:The CT images got are pre-processed, obtain the bony areas in the CT images;
Wherein, the bony areas may refer to shown in Fig. 5, wherein, the region marked as 1 in described Fig. 5 can generation
Table bony areas, can be by bony areas labeled as pre-set color, for example, by Fig. 5 in order to distinguish bony areas and other regions
Zone marker marked as 1 is red.
Step S103:Class elliptic region positioning is carried out to obtained each bony areas, obtains being imaged as class ellipse
The bony areas in face;
Referring to Fig. 2, based on synosteology structure, particularly ball and socket joint almost spherical feature, it is in the axial face of bone
Reveal the elliptical shape of class.It, can be adjacent by detecting the class ellipsoid realization referring to the bony areas for being is marked in Fig. 6
Bony areas is split as multiple independent sub- bony areas by the differentiation between bone, and each independent sub- bony areas is
The bony areas being made of one piece of complete bone;
Step S104:It is selected from obtained each class ellipsoid using Hough transform foundation preset rules at least one
Class ellipsoid, by the class ellipsoid of selection labeled as seed region, referring to Fig. 6, the seed region in Fig. 6 is the area labeled as 2
Domain in order to which seed region and other bony areas are distinguished, can be used and the seed region is carried out using special marking
Seed region marked as 2 is labeled as green by label, such as Fig. 6, and red is labeled as labeled as 1 bony areas;
In view of femoral head or ilium the elliptical characteristic of class can be also showed in axial face some regions, it is therefore desirable to from
Specific class elliptic region is selected in all elliptic regions detected as seed region, so that the seed that positioning obtains
Region is the corresponding bony areas of target bone, and when needing to select different bones as target area, it is right with it only to select
The preset rules answered.
In the class elliptic region in obtaining each CT images, Hough transform can be used and go to detect in all CT images
Class ellipsoid.It is contemplated that the difference of sample bone difference and CT equipment, directly with Hough transform, it is difficult to accurately examine
Measure class ellipsoid.Especially when bone is in the case where there is lesion, accuracy rate becomes lower.Therefore this method is proposed with wide
Adopted Hough transform goes to detect such elliptic region.Using Generalized Hough Transform on the one hand in detection speed than directly using
Hough transform is faster, the occurrence of on the other hand can also effectively reducing missing inspection.
Step S105:It is background area by non-bone zone marker by the seed region labeled as target area;
In this step, referring to Fig. 6, emphasis label is carried out using the seed region as the target area to be observed, with
So that it has significant difference with other regions;
Step S106:Referring to Fig. 7, using the partitioning algorithm based on graph theory, target area and background by above-mentioned setting
Region separates the bony areas for being connected to one piece in joint, and different marks is set to different bony areas;
In this step, after label obtains the target area, the CT is preferably individually observed in order to facilitate user
Each bony areas being made of independent bone in image needs to separate each adjacent bony areas in this step
Come, this step is split the bony areas in image, will be connected to by cutting algorithm (GraphCuts) based on graph theory
One piece of bony areas is separated in joint.GraphCuts algorithms are the gray value and coordinate using CT images, and structure one has
Vertex and the collection of illustrative plates on side set area-of-interest and background area, using the principle of energy minimum, energy minimization path are made
For boundary, collection of illustrative plates is divided into the process of area-of-interest and background area.
Referring to Fig. 8, when corresponding bony areas (seed region) label of a targeted bone in the CT images is completed
Afterwards, method disclosed in performing the above embodiments of the present application is repeated, is continued to the corresponding seed region of bones other in CT images
It is marked, certainly, the differentiation of different seed regions for convenience, the mark of each seed region is different, such as in Fig. 8, kind
Subregion 2 uses red-label, and seed region 3 uses light blue color marker, and seed region 4 uses Green Marker, and seed region 5 is adopted
With dark blue color marker, and use referring to Fig. 9 the segmentation result effect of each seed region of Surface Rendering Three-dimensional Displays
Figure.
In above-mentioned technical proposal disclosed in the embodiment of the present application, the pretreatment can include
A1:Positioned to obtain the human region in image to the CT images got;
A2:Noise reduction process is carried out to the human region that positioning obtains;
A3:Bony areas positioning is carried out to the human region after noise reduction process and removes soft tissue area, obtains the CT
Bony areas in image.
Wherein, when positioning human region, process can be:
By locating points of interest in CT images, each connected region of point of interest in CT images is obtained, and retention volume is maximum
Connected region;
This step is mainly to remove other interference regions such as bed body in CT images.This step can specifically pass through threshold method
By positioning interest region in CT images.
The noise reduction process can be completed by Gaussian filter;
Bony areas positioning is carried out to the human region after noise reduction process and removes the detailed process of soft tissue area can be with
Including:
Mark to obtain the bony areas in image according to threshold method;
By calcified tissue, process can be specially for removal:Using morphological operator removal label obtain by calcified tissue
The bony areas of formation.In this process, by judge whether with the non-conterminous bony areas of other bony areas, if
In the presence of, then it is believed that the bony areas be calcified tissue, cancel the label to its bony areas.
Soft tissue is removed, process is specifically as follows:Bony areas after removal calcified tissue is handled, obtains institute
State the image greyscale histogram of bony areas;Gray value in image greyscale histogram is less than to the gray scale of the bony areas of preset value
Value is set as in the example of 0, Figure 10 the preset value being set as 80, i.e., by gray value in image greyscale histogram less than 80
The gray value of bony areas is set as 0, during actual use, the size of the preset value can voluntarily be set according to user demand
It puts.
It is described to select at least one class oval from obtained each class ellipsoid according to preset rules in the above method
Face, and labeled as seed region, can specifically include:
According to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6The score value S of each class ellipsoid is calculated, will divide
The highest class ellipsoids of value S are labeled as seed region;The formula can be by being instructed using machine learning algorithm [such as BP neural network]
It gets;
Wherein, the x1For the edge gray scale of class ellipsoid and center gray value ratio, the x2Major and minor axis for class ellipsoid
Than x3For the area of class ellipsoid, x4For the long axis of class ellipsoid and the angle of horizontal plane, x5For class ellipsoid central point with
The distance of target area central point, the target area are according to the bony areas in threshold method labeling CT image, x6It is ellipse for class
The boundary of disc and the distance of soft tissue boundary line, the soft tissue boundary line is that gray value is in the image greyscale histogram
The boundary that the position of preset value is formed in the bony areas;The a1、a2、a3、a4、a5、a6Respectively predetermined coefficient, foundation
The difference of the bone of required positioning, a1、a2、a3、a4、a5、a6It is of different sizes, wherein, a1、a2、a3、a4、a5、a6's
Value can according to BP neural network algorithm using corresponding bone training sample set and test sample set to formula S=
a1x1+a2x2+a3x3+a4x4+a5x5+a6x6It is trained to obtain.For example, the bone when required positioning (need to close hip for hip joint
Corresponding bony areas is saved as target area) when, according to BP neural network algorithm using the training sample set of hip joint and
Test sample set is trained above-mentioned formula to obtain a1、a2、a3、a4、a5、a6Value, a can be obtained1Equal to 0.3, a2It is equal to
0.2, a3Equal to 0.2, a4Equal to 0.15, a5Equal to 0.1, a6Equal to 0.05.
In order to further improve the accuracy that above-mentioned formula calculates structure in actual use, may be used also in the above method
To include:Training sample input by user is obtained, the training sample is put into the training sample set of corresponding bone, according to
According to updated training sample set and test sample set to formula S=a corresponding to the bone1x1+a2x2+a3x3+a4x4+
a5x5+a6x6It is trained, according to a in training result update above-mentioned formula1、a2、a3、a4、a5、a6Value, wherein, the use
The training sample of family input can be the training sample of the corresponding highest class ellipsoids of score value S, specially:It is defeated to obtain user
The true score value S ' of the highest class ellipsoids of score value S entered, the true score value S ' can be by experts in the CT images of output
Class ellipsoid scored to obtain, by the true score value S ' and its corresponding x1、x2、x3、x4、x5And x6Import training sample
This set, according to training sample set again to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6It is trained, according to instruction
Practice result and adjust a1、a2、a3、a4、a5And a6Size.
Corresponding to the above method, disclosed herein as well is a kind of Bones and joints CT image segmentation systems, referring to Figure 11, the system
It can include:
Pretreatment unit 100, is used for:To the CT images got, the CT images are pre-processed, obtain the CT
Bony areas in image;
Class elliptic region positioning unit 200, for carrying out class elliptic region positioning to obtained each bony areas,
Obtain being imaged as the bony areas of class ellipsoid;
Seed region positioning unit 300, for selecting at least one from obtained each class ellipsoid according to preset rules
The class ellipsoid of selection is labeled as seed region by a class ellipsoid;
Zone marker unit 400, for the bony areas corresponding to the seed region labeled as target area, will to be removed
Zone marker other than the bony areas is background area.
Unit 500 is identified, for utilizing the partitioning algorithm based on graph theory, target area and background area by above-mentioned setting
Domain separates the bony areas for being connected to one piece in joint, and different marks is set to different bony areas.
Wherein, the pretreatment unit 100 can include, human region extraction unit, for the CT images got
Positioned to obtain the human region in image;Noise reduction unit, the human region for being obtained to positioning carry out noise reduction process;Bone
Bone area extracting unit for carrying out bony areas positioning to the human region after noise reduction process and removing soft tissue area, obtains
To the bony areas in the CT images.
Corresponding with the above method, above system can also include boundary and optimize unit, for setting different marks
Bony areas into row bound optimize.
In order to further improve the accuracy that above-mentioned formula calculates structure in actual use, in above system, also
Including:
Formula optimization unit is inside configured with training sample database and test sample database, the number of training
It is used to store the training sample of the advance data of user according to library, test sample database is used for the test specimens that training user pre-enters
This, the formula optimization unit can also put the parameter of the highest class ellipsoids of the score value S obtained every time as training sample
Enter the training sample database, certainly, due to the score value S that is calculated by formula can there are error, be put into training
The true score value S ' of the training sample of sample database is input by user, when the training sample database adds new training
After sample, the formula optimization unit 140 is based on BT algorithms automatically according to the true score value of the highest class ellipsoids of the score value S
S’、x1、x2、x3、x4、x5And x6To formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6It is trained, according to training result tune
The whole a1、a2、a3、a4、a5And a6Size.
For convenience of description, it is divided into various modules during description system above with function to describe respectively.Certainly, implementing this
The function of each module is realized can in the same or multiple software and or hardware during application.
Each embodiment in this specification is described by the way of progressive, identical similar portion between each embodiment
Point just to refer each other, and the highlights of each of the examples are difference from other examples.Especially for system or
For system embodiment, since it is substantially similar to embodiment of the method, so describing fairly simple, related part is referring to method
The part explanation of embodiment.System and system embodiment described above is only schematical, wherein the conduct
The unit that separating component illustrates may or may not be it is physically separate, the component shown as unit can be or
Person may not be physical unit, you can be located at a place or can also be distributed in multiple network element.It can root
Factually border needs to select some or all of module therein realize the purpose of this embodiment scheme.Ordinary skill
Personnel are without creative efforts, you can to understand and implement.
Professional further appreciates that, with reference to each exemplary unit of the embodiments described herein description
And algorithm steps, can be realized with the combination of electronic hardware, computer software or the two, in order to clearly demonstrate hardware and
The interchangeability of software generally describes each exemplary composition and step according to function in the above description.These
Function is performed actually with hardware or software mode, specific application and design constraint depending on technical solution.Profession
Technical staff can realize described function to each specific application using distinct methods, but this realization should not
Think beyond the scope of this invention.
It can directly be held with reference to the step of method or algorithm that the embodiments described herein describes with hardware, processor
The combination of capable software module or the two is implemented.Software module can be placed in random access memory (RAM), memory, read-only deposit
Reservoir (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technology
In any other form of storage medium well known in field.
It should also be noted that, herein, relational terms such as first and second and the like are used merely to one
Entity or operation are distinguished with another entity or operation, without necessarily requiring or implying between these entities or operation
There are any actual relationship or orders.Moreover, term " comprising ", "comprising" or its any other variant are intended to contain
Lid non-exclusive inclusion, so that process, method, article or equipment including a series of elements not only will including those
Element, but also including other elements that are not explicitly listed or further include as this process, method, article or equipment
Intrinsic element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that
Also there are other identical elements in process, method, article or equipment including the element.
The foregoing description of the disclosed embodiments enables professional and technical personnel in the field to realize or use the present invention.
A variety of modifications of these embodiments will be apparent for those skilled in the art, it is as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, it is of the invention
The embodiments shown herein is not intended to be limited to, and is to fit to and the principles and novel features disclosed herein phase one
The most wide range caused.
Claims (10)
1. a kind of Bones and joints CT image partition methods, which is characterized in that including:
To the CT images got, the CT images are pre-processed, obtain the bony areas in the CT images;
Class elliptic region positioning is carried out to obtained each bony areas using Hough transform, obtains being imaged as class ellipse
The bony areas in face;
At least one class ellipsoid is selected from obtained each class ellipsoid according to preset rules, by the class ellipsoid mark of selection
It is denoted as seed region;
It is background area by non-bone zone marker by the seed region labeled as target area;
Using the energy smallest partition algorithm based on graph theory, by the target area and background area of above-mentioned label, will be connected to
One piece of bony areas is separated in joint, and different marks is set to different bony areas.
2. Bones and joints CT image partition methods according to claim 1, the pretreatment includes:Including:
The CT images got are positioned, obtain the human region in image;
Noise reduction process is carried out to the human region that positioning obtains;
Bony areas positioning is carried out to the human region after noise reduction process and removes soft tissue area, is obtained in the CT images
Bony areas.
3. CT image partition methods according to claim 1, which is characterized in that closing the bony areas for being connected to one piece
It separates at section, and after the mark different to different bony areas settings, further includes:
The boundary of different bony areas is optimized.
4. CT image partition methods according to claim 1, which is characterized in that described each by what is obtained according to preset rules
At least one class ellipsoid is selected in a class ellipsoid, including:
According to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6The score value S of each class ellipsoid is calculated, by score value S most
High one or more class ellipsoids are labeled as seed region;
Wherein, the x1For the edge gray scale of class ellipsoid and center gray value ratio, the x2For the axial ratio of class ellipsoid,
x3For the area of class ellipsoid, x4For the long axis of class ellipsoid and the angle of horizontal plane, x5Central point and target for class ellipsoid
The distance of regional center, the target area are according to the bony areas in binary-state threshold method labeling CT image, x6It is ellipse for class
The boundary of disc and the distance of soft tissue boundary line, the soft tissue boundary line is that gray value is in the image greyscale histogram
The boundary that the position of preset value is formed in the bony areas;The a1、a2、a3、a4、a5、a6Respectively predetermined coefficient.
5. CT image partition methods according to claim 4, which is characterized in that further include:Obtain score value S input by user
The true score value S ' of highest class ellipsoid, according to the true score value S ', x1、x2、x3、x4、x5And x6To formula S=a1x1+
a2x2+a3x3+a4x4+a5x5+a6x6It is trained, a is adjusted according to training result1、a2、a3、a4、a5And a6Size.
6. a kind of Bones and joints CT image segmentation systems, which is characterized in that including:
Pretreatment unit, for the CT images got, pre-processing, being obtained in the CT images to the CT images
Bony areas;
Class elliptic region positioning unit, for carrying out class area elliptica to obtained each bony areas using Hough transform
Domain positions, and obtains being imaged as the bony areas of class ellipsoid;
Seed region positioning unit, for according to preset rules by selecting at least one class oval in obtained each class ellipsoid
The class ellipsoid of selection is labeled as seed region by face;
Indexing unit, for labeled as target area, the bony areas corresponding to the seed region will to be removed the bone area
Zone marker other than domain is background area;
Indexing unit, for using utilizing the energy smallest partition algorithm based on graph theory, by the target area of above-mentioned label and
Background area separates the bony areas for being connected to one piece in joint, and different marks is set to different bony areas.
7. Bones and joints CT image segmentation systems according to claim 6, which is characterized in that the specific quilt of the pretreatment unit
It is configured to:
Position the human region in the CT images got;Noise reduction process is carried out to the human region that positioning obtains;At noise reduction
Human region after reason carries out bony areas positioning and removes soft tissue area, obtains the bony areas in the CT images.
8. Bones and joints CT image segmentation systems according to claim 6, which is characterized in that further include:
Optimize unit, for being optimized to the bony areas for setting different marks into row bound.
9. Bones and joints CT image segmentation systems according to claim 6, which is characterized in that the seed region positioning unit
It is specifically configured to:
According to formula S=a1x1+a2x2+a3x3+a4x4+a5x5+a6x6The score value S of each class ellipsoid is calculated, by score value S most
High one or more class ellipsoids are labeled as seed region;
Wherein, the x1For the edge gray scale of class ellipsoid and center gray value ratio, the x2For the axial ratio of class ellipsoid,
x3For the area of class ellipsoid, x4For the long axis of class ellipsoid and the angle of horizontal plane, x5Central point and target for class ellipsoid
The distance of regional center, the target area are according to the bony areas in binary-state threshold method labeling CT image, x6It is ellipse for class
The boundary of disc and the distance of soft tissue boundary line, the soft tissue boundary line is that gray value is in the image greyscale histogram
The boundary that the position of preset value is formed in the bony areas;The a1、a2、a3、a4、a5、a6Respectively predetermined coefficient.
10. Bones and joints CT image segmentation systems according to claim 9, which is characterized in that further include:
Formula optimization unit, for obtaining the true score value S ' of the highest class ellipsoids of score value S input by user, according to described point
True score value S ', the x of the highest class ellipsoids of value S1、x2、x3、x4、x5And x6To formula S=a1x1+a2x2+a3x3+a4x4+a5x5+
a6x6It is trained, a is adjusted according to training result1、a2、a3、a4、a5And a6Size.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611263197.3A CN108269261A (en) | 2016-12-30 | 2016-12-30 | A kind of Bones and joints CT image partition methods and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201611263197.3A CN108269261A (en) | 2016-12-30 | 2016-12-30 | A kind of Bones and joints CT image partition methods and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108269261A true CN108269261A (en) | 2018-07-10 |
Family
ID=62755377
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201611263197.3A Pending CN108269261A (en) | 2016-12-30 | 2016-12-30 | A kind of Bones and joints CT image partition methods and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108269261A (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109191475A (en) * | 2018-09-07 | 2019-01-11 | 博志科技有限公司 | Terminal plate of vertebral body dividing method, device and computer readable storage medium |
CN109192267A (en) * | 2018-08-09 | 2019-01-11 | 深圳狗尾草智能科技有限公司 | Virtual robot is accompanied in movement |
CN109741360A (en) * | 2019-01-07 | 2019-05-10 | 上海联影医疗科技有限公司 | A kind of Bones and joints dividing method, device, terminal and readable medium |
CN109919935A (en) * | 2019-03-12 | 2019-06-21 | 语坤(北京)网络科技有限公司 | A kind of neck body blood vessel segmentation method and apparatus |
CN111105414A (en) * | 2019-12-31 | 2020-05-05 | 杭州依图医疗技术有限公司 | Processing method, interaction method, display method and storage medium |
CN112017146A (en) * | 2019-05-31 | 2020-12-01 | 杭州三坛医疗科技有限公司 | Skeleton segmentation method and device |
CN114037719A (en) * | 2022-01-11 | 2022-02-11 | 北京威高智慧科技有限公司 | Bone region determination method and device, electronic equipment and storage medium |
CN114066885A (en) * | 2022-01-11 | 2022-02-18 | 北京威高智慧科技有限公司 | Lower limb skeleton model construction method and device, electronic equipment and storage medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8649577B1 (en) * | 2008-11-30 | 2014-02-11 | Image Analysis, Inc. | Automatic method and system for measurements of bone density and structure of the hip from 3-D X-ray imaging devices |
CN104657984A (en) * | 2015-01-28 | 2015-05-27 | 复旦大学 | Automatic extraction method of three-dimensional breast full-volume image regions of interest |
CN105590092A (en) * | 2015-11-11 | 2016-05-18 | 中国银联股份有限公司 | Method and device for identifying pupil in image |
US20160275674A1 (en) * | 2015-12-29 | 2016-09-22 | Laboratoires Bodycad Inc. | Method and system for performing multi-bone segmentation in imaging data |
-
2016
- 2016-12-30 CN CN201611263197.3A patent/CN108269261A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8649577B1 (en) * | 2008-11-30 | 2014-02-11 | Image Analysis, Inc. | Automatic method and system for measurements of bone density and structure of the hip from 3-D X-ray imaging devices |
CN104657984A (en) * | 2015-01-28 | 2015-05-27 | 复旦大学 | Automatic extraction method of three-dimensional breast full-volume image regions of interest |
CN105590092A (en) * | 2015-11-11 | 2016-05-18 | 中国银联股份有限公司 | Method and device for identifying pupil in image |
US20160275674A1 (en) * | 2015-12-29 | 2016-09-22 | Laboratoires Bodycad Inc. | Method and system for performing multi-bone segmentation in imaging data |
Non-Patent Citations (3)
Title |
---|
周生俊: "医学CT图像分割方法研究", 《中国博士学位论文全文数据库信息科技辑》 * |
邓勃: "《挑战人脑 计算机在化学中的应用》", 30 September 1998, 湖南教育出版社 * |
韦轶群等: "基于自适应Graph Cuts的自动股骨头分割", 《上海交通大学学报》 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109192267A (en) * | 2018-08-09 | 2019-01-11 | 深圳狗尾草智能科技有限公司 | Virtual robot is accompanied in movement |
CN109191475A (en) * | 2018-09-07 | 2019-01-11 | 博志科技有限公司 | Terminal plate of vertebral body dividing method, device and computer readable storage medium |
CN109741360A (en) * | 2019-01-07 | 2019-05-10 | 上海联影医疗科技有限公司 | A kind of Bones and joints dividing method, device, terminal and readable medium |
CN109741360B (en) * | 2019-01-07 | 2022-02-22 | 上海联影医疗科技股份有限公司 | Bone joint segmentation method, device, terminal and readable medium |
CN109919935A (en) * | 2019-03-12 | 2019-06-21 | 语坤(北京)网络科技有限公司 | A kind of neck body blood vessel segmentation method and apparatus |
CN109919935B (en) * | 2019-03-12 | 2021-02-09 | 语坤(北京)网络科技有限公司 | Head and neck blood vessel segmentation method and device |
CN112017146A (en) * | 2019-05-31 | 2020-12-01 | 杭州三坛医疗科技有限公司 | Skeleton segmentation method and device |
CN112017146B (en) * | 2019-05-31 | 2024-03-22 | 杭州三坛医疗科技有限公司 | Bone segmentation method and device |
CN111105414A (en) * | 2019-12-31 | 2020-05-05 | 杭州依图医疗技术有限公司 | Processing method, interaction method, display method and storage medium |
CN114037719A (en) * | 2022-01-11 | 2022-02-11 | 北京威高智慧科技有限公司 | Bone region determination method and device, electronic equipment and storage medium |
CN114066885A (en) * | 2022-01-11 | 2022-02-18 | 北京威高智慧科技有限公司 | Lower limb skeleton model construction method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108269261A (en) | A kind of Bones and joints CT image partition methods and system | |
CN105913432B (en) | Aorta extracting method and device based on CT sequence images | |
JP5755734B2 (en) | Method and system for automatically determining geometric elements from 3D medical images of bone | |
CN104809740B (en) | Knee cartilage image automatic segmentation method based on SVM and Hookean region growth | |
US9514533B2 (en) | Method for determining bone resection on a deformed bone surface from few parameters | |
CN111161241B (en) | Liver image identification method, electronic equipment and storage medium | |
CN107292928A (en) | A kind of method and device of blood vessel positioning | |
CN104851107A (en) | Vertebra positioning method based on CT sequence image | |
WO2005009240A9 (en) | Image processing method and computer-readable recording medium containing image processing program | |
CN106127753B (en) | CT images body surface handmarking's extraction method in a kind of surgical operation | |
CN103294883A (en) | Method and system for intervention planning for transcatheter aortic valve implantation | |
CN105139377A (en) | Rapid robustness auto-partitioning method for abdomen computed tomography (CT) sequence image of liver | |
CN114261095B (en) | AI-based orthopedic 3D printing method and device | |
CN114642444A (en) | Oral implantation precision evaluation method and system and terminal equipment | |
CN106373138B (en) | For extracting the image processing method and device of dental arch curve | |
Zou et al. | Semi-automatic segmentation of femur based on harmonic barrier | |
CN113469935B (en) | Automatic detection and positioning method for posterior superior iliac spine based on CT image | |
CN110766667A (en) | Damaged bone repair modeling method based on contour sample library matching and template guidance | |
CN109816672A (en) | A kind of image segmentation acquisition methods and device for neck body bone | |
Yao et al. | Quantitative vertebral compression fracture evaluation using a height compass | |
CN116029972A (en) | Fracture region nondestructive segmentation and reconstruction method based on morphology | |
CN105678738B (en) | The localization method and its device of datum mark in medical image | |
Subburaj et al. | 3D shape reasoning for identifying anatomical landmarks | |
CN114581395A (en) | Method for detecting key points of spine medical image based on deep learning | |
Gu et al. | 3D automatic fiducial marker localization approach for frameless stereotactic neuro-surgery navigation |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180710 |
|
RJ01 | Rejection of invention patent application after publication |