CN101807296A - Segmentation method of medical ultrasonic image three-dimensional target object - Google Patents
Segmentation method of medical ultrasonic image three-dimensional target object Download PDFInfo
- Publication number
- CN101807296A CN101807296A CN200910005673A CN200910005673A CN101807296A CN 101807296 A CN101807296 A CN 101807296A CN 200910005673 A CN200910005673 A CN 200910005673A CN 200910005673 A CN200910005673 A CN 200910005673A CN 101807296 A CN101807296 A CN 101807296A
- Authority
- CN
- China
- Prior art keywords
- node
- ultrasonic image
- medical ultrasonic
- destination object
- snake model
- 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
Images
Landscapes
- Ultra Sonic Daignosis Equipment (AREA)
Abstract
The embodiment of the invention discloses a segmentation method of a medical ultrasonic image three-dimensional target object, which is invented for solving the problems of low segmentation precision or nonideal time performance or low automation degree in the prior art. The segmentation method comprises that initial scattered medical ultrasonic image data including the medical ultrasonic image three-dimensional target object is restructured and a regular data set is obtained and segmented on the basis of a contour line. The concepts of keyframe and middle frame are led into the segmentation based on the contour line. A topological self-adaption deformation model and an interpolation based on shape are applied. The contour of the medical ultrasonic image three-dimensional target object can be obtained quickly and accurately. The invention is applicable to not only the segmentation of the medical ultrasonic image three-dimensional target object but also the segmentation of CT, MRI and other medical image order three-dimensional target objects.
Description
Technical field
The present invention relates to medical ultrasonic image analysis and processing technology field, relate in particular to a kind of dividing method of medical ultrasonic image three-dimensional destination object.
Background technology
In medical ultrasonic image analysis and processing technology field, cutting apart of objective object has important clinical application value.But because the constraint of medical ultrasound image mechanism, technical performance indexs such as the signal to noise ratio (S/N ratio) of image, spatial resolution and contrast are not ideal enough, and cutting apart of feasible wherein objective object is comparatively difficult.Existing dividing method mainly contains the method for manual demarcation of boundary, based on the method for region growing with based on method of traditional activity skeleton pattern etc.
Wherein, the dividing method of manual demarcation of boundary needs operating personnel to depict the edge of destination object to be split in cutting procedure in every two field picture, all need manually-operated in whole process, lack robotization, it is bigger that testing result is influenced by operating personnel's subjective factor.Dividing method based on region growing must be provided with seed points, and algorithm execution speed is slow, lacks general convergence criterion, and choosing of seed points and convergence criterion has very big influence to the result.Dividing method based on the traditional activity skeleton pattern can be realized the objective Object Segmentation under less manual intervention, but this method to the processing power of complex topology structure a little less than, performance index such as segmentation effect and segmentation precision are difficult to satisfy the clinical practice demand.
In realizing process of the present invention, the inventor finds that there are the following problems at least in the prior art:
The dividing method of existing medical ultrasonic image three-dimensional destination object, segmentation precision and automaticity are lower, and time performance is relatively poor.
Summary of the invention
Embodiments of the invention provide a kind of dividing method of medical ultrasonic image three-dimensional destination object, have higher segmentation precision and automaticity, and time performance is good.
For achieving the above object, embodiments of the invention adopt following technical scheme:
A kind of dividing method of medical ultrasonic image three-dimensional destination object comprises:
To the rearrangement of recombinating of original view data at random, obtain the rule body data set, described original view data at random is the primitive medicine ultrasonic image data that comprise described medical ultrasonic image three-dimensional destination object;
Described rule body data set is carried out cutting apart based on outline line.
The dividing method of the medical ultrasonic image three-dimensional destination object that the embodiment of the invention provides, at first to the primitive medicine ultrasonic image data that the comprise described medical ultrasonic image three-dimensional destination object rearrangement of recombinating, obtain the rule body data set, then described rule body data set is carried out cutting apart based on outline line.Introduced the notion of key frame and intermediate frame described based in the cutting apart of outline line, used topological self-adaptation deformation model and, can obtain the profile of described medical ultrasonic image three-dimensional destination object quickly and accurately based on the interpolation of shape.Compared with prior art, whole cutting procedure has higher segmentation precision and certain automaticity, and has time performance preferably.
Description of drawings
The dividing method process flow diagram of the medical ultrasonic image three-dimensional destination object that Fig. 1 provides for the embodiment of the invention;
Definite synoptic diagram of T-Snake model node external force normal vector direction in the dividing method of the medical ultrasonic image three-dimensional destination object that Fig. 2 provides for the embodiment of the invention;
Definite synoptic diagram of T-Snake model node external force normal vector direction in the dividing method of the medical ultrasonic image three-dimensional destination object that Fig. 3 provides for the embodiment of the invention;
T-Snake model node location upgrades synoptic diagram in the dividing method of the medical ultrasonic image three-dimensional destination object that Fig. 4 provides for the embodiment of the invention;
Node synoptic diagram after upgrading in the dividing method of the medical ultrasonic image three-dimensional destination object that Fig. 5 provides for the embodiment of the invention.
Embodiment
Embodiments of the invention provide a kind of dividing method of medical ultrasonic image three-dimensional destination object.
For the advantage that makes technical solution of the present invention is clearer, the present invention is elaborated below in conjunction with drawings and Examples.
As shown in Figure 1, the dividing method of described medical ultrasonic image three-dimensional destination object comprises:
S101, to the rearrangement of recombinating of original view data at random, obtain the rule body data set.
Described original view data at random is the primitive medicine ultrasonic image data that comprise described medical ultrasonic image three-dimensional destination object, and the relative space position of described primitive medicine ultrasonic image data does not have rule, is at random, irregular data set.Reset to handle by described original view data at random is recombinated, make the described primitive medicine ultrasonic image data that comprise the objective object be filled on the three dimensional network lattice point of the rule that square forms the formation rule volumetric data set.In the present embodiment, described reorganization is reset to handle and is specially the interpolation smoothing processing, and wherein, described interpolation smoothing processing comprises that voxel faces recently that the territory interpolation is level and smooth, pixel is faced methods such as interpolation level and smooth and distance weighted interpolation in territory is level and smooth recently.
S102, select some key frames in described rule body data centralization.
Wherein, the rule body data centralization that in step S101, obtains, extract equally spaced two-dimentional parallel image sequence frame, human tissue organ's contour shape on each two-dimentional parallel image sequence frame is to change continuously rather than sudden change, described two-dimentional parallel image sequence frame comprises more than one sequence frame section, and initial and end two frames of each sequence frame section are key frame.Wherein, described key frame can be selected further combined with doctor's the clinical experience and the anatomical features of human tissue organ.
S103, on described key frame, extract the profile of described medical ultrasonic image three-dimensional destination object based on T-Snake model (Topologically Adaptable Snakes, topological self-adaptation deformation model).Here, described T-Snake model is a kind of improvement to traditional T-Snake model.
The specific implementation process of the described profile that extracts described medical ultrasonic image three-dimensional destination object on described key frame is as follows:
S103a, carry out the triangular grids of image space; The described T-Snake model of initialization is a closed curve with arbitrary shape, and in the present embodiment, described T-Snake model is initialized as a circle, and the intersection point of described T-Snake model and above-mentioned triangular mesh is the start node of described T-Snake model.
Wherein, described T-Snake model is a deformation model that topology is variable, and is stronger for the processing and the adaptive faculty of complex topology structure, is defined as the closed contour that is formed by connecting by N node, described N is a positive integer, and N 〉=2, wherein, i (i=0 ..., N-1) equation of motion of individual node is:
In the formula,
Be the function of position of i node,
Be the movement velocity of i node, γ
iBe ratio of damping.
Wherein, two internal energies that are referred to as described T-Snake model comprise tensile force behind the equation left side
And bending force
Wherein,
Be the tensile force at i node place,
It is the bending force at i node place; Parameter a is controlling the stretching intensity of force, and parameter b is being controlled crooked intensity of force, and internal energy is the flatness constraint condition of model.Two external energies that are referred to as described T-Snake model in equation the right comprise bulging force
With external force based on image gradient features
Bulging force
Direction be called the normal vector direction of the external force of node, be the moving direction of the node of described T-Snake model; External force based on image gradient features
Can make model converge on efficient frontier.The motion of supposing node does not have inertia, then when the internal energy of T-Snake model and external energy reach equilibrium state, and the movement velocity vanishing of node, and the node that is in the energy equilibrium state is the point in the image.
There are some shortcomings in traditional T-Snake model when using: in model deformation evolutionary process, if the spacing of adjacent three nodes is bigger, carry out to produce bigger error when differential calculation is asked internal force, and node selfing phenomenon can occur, cause model to converge on incorrect edge or the profile; And the direction of external force normal vector can directly have influence on the deformation evolution situation of traditional T-Snake model, also can cause model to converge on incorrect edge or the profile under the extreme case.
In order to ensure the accuracy based on the objective Object Segmentation of outline line, the embodiment of the invention has proposed a kind of improved T-Snake model, and described improvement comprises:
(1) calculating of internal force
In the internal energy structure of tradition T-Snake model, the computing formula of tensile force and bending force is as follows respectively:
In the formula,
Be the function of position of i node,
Being the tensile force at i node place, is i node location function
The discrete approximation of second derivative;
Being the bending force at i node place, is i node location function
The discrete approximation of quadravalence derivative.
Know that by the computing method theory distance that the demanding differential approximate treatment of numerical precision needs Centroid to be adjacent node is answered approximately equal; And, internodal distance should be made normalized for unequal situation, therefore, the embodiment of the invention provides the computing formula of improved tensile force and bending force:
In the formula,
Be the function of position of i node,
Be the tensile force at i node place,
It is the bending force at i node place; d
1Be the distance between i node and i-1 the node, d
2It is the distance between i+1 node and i the node.
(2) external force normal vector direction determines
The definite of the normal vector direction of the external force of described each node of T-Snake model carries out according to following steps:
Through a node in the described T-Snake model and two other node of its next-door neighbour, determine a circle, wherein, described three nodes are not on same straight line; If the center of circle of described circle is in the border of described T-Snake model, the normal vector direction of the external force of this node is that radial finger along above-mentioned circle is to this node; If the center of circle of above-mentioned circle is outside the border of described T-Snake model, the normal vector direction of the external force of this node is for radially deviating from this node along above-mentioned circle.
As Fig. 2, shown in Figure 3, quadrilateral (curve 2) is the border of T-Snake model, node a is a node on the described T-Snake model boundary, node b, c be on the described T-Snake model boundary, with two other node of node a next-door neighbour, described node a, b, c be not on same straight line, through a, b, 3 determined circles of c is curve 1, and the center of circle is o.As shown in Figure 2, because o is in the border of T-Snake model, the external force normal vector direction at b point place then is defined as among Fig. 2
Shown in, to reach the purpose of model towards true profile 3 directions expansion.As shown in Figure 3, because center of circle o is outside the border of T-Snake model, the external force normal vector direction at b point place then is defined as among Fig. 3
Shown in, to reach the purpose of model towards true profile 3 directions expansion.
Adopt the computing formula of above-mentioned T-Snake model internal force, not only can improve the numerical evaluation precision of method, and, can avoid model node selfing phenomenon further according to definite method of described external force normal vector direction.
S103b, calculate the internal force and the external force of described each node of T-Snake model, determine the normal vector direction of the external force of each node, judge whether the internal force and the external force of each node of described T-Snake model all reaches balance.The definite of the normal vector direction of the calculating of described internal force and the external force of each node carries out according to the described method of step S103a.
S103c, if the internal force and the external force of each node of described T-Snake model all do not reach balance, then the node location in the described T-Snake model is upgraded, the normal vector direction of each node along the external force of this node moved, obtain the described T-Snake model after the deformation.As shown in Figure 4, internal layer is an origin node, and skin is the node after the position renewal.
S103d, as shown in Figure 5 redefines the described T-Snake model after the deformation and the intersection point of triangular mesh, with it as the new node of described T-Snake model.
Whether internal force and the external force of judging each node of the T-Snake model after the described deformation all reach balance, if all do not reach balance, repeating step S103b-S103d then, till the internal force of each node of T-Snake model after the described deformation and external force all reach equilibrium state, then described profile at the described medical ultrasonic image three-dimensional destination object that extracts on the described key frame is: all nodes of described T-Snake model under the described equilibrium state, the matched curve that obtains by B-spline-fitting or other similar disposal routes.
Because the number of selected key frame is less usually, calculated amount is very little, and this has just guaranteed that the method for the invention has time performance preferably.
S104, by on described intermediate frame, extracting the profile of described medical ultrasonic image three-dimensional destination object based on the interpolation method of shape, wherein, the picture frame between the described key frame is an intermediate frame.
The specific implementation process of described interpolation method is as follows:
S104a, the profile that obtains according to step S103, described key frame is carried out range conversion, obtain key frame distance figure, on behalf of this, the absolute value of each point put the minimum distance of profile on the distance map, and value is greater than Regional Representative's profile outside of zero, minus Regional Representative's profile inside.
S104b, the described key frame distance figure of two adjacent key frames is carried out linear interpolation, obtain the intermediate frame distance map.Wherein, described linear interpolation is a unknown data of estimating the 3rd correspondence according to known two relevant corresponding relation data, and in the present embodiment, known two is two adjacent key frame distance figure, and the 3rd is the intermediate frame distance map.
At zero point on S104c, the described intermediate frame distance map of extraction, form the profile of the described medical ultrasonic image three-dimensional destination object on this intermediate frame described each zero point.
In the present embodiment, the step that form the profile of the described medical ultrasonic image three-dimensional destination object on this intermediate frame described each zero point is specially: connected described each zero point successively, form an open curve, described open curve i.e. the profile of the described medical ultrasonic image three-dimensional destination object on this intermediate frame.
Repeating step S104b-S104c, all pass through linear interpolation up to the key frame distance figure of any two adjacent key frames and obtain the intermediate frame distance map, extract the zero point on all intermediate frame distance maps, connected any two adjacent zero points, form a closed curve, described closed curve is the profile of the described medical ultrasonic image three-dimensional destination object on all intermediate frames.
Because the calculated amount of described interpolation method is very little, this has just further guaranteed that the method for the invention has time performance preferably.
The dividing method of the medical ultrasonic image three-dimensional destination object that the embodiment of the invention provides, at first to the primitive medicine ultrasonic image data that the comprise described medical ultrasonic image three-dimensional destination object rearrangement of recombinating, obtain the rule body data set, then based on the outline line of improved T-Snake model at the described medical ultrasonic image three-dimensional destination object of key-frame extraction, by intermediate frame being carried out on whole rule body data sets, extract the outline line of described medical ultrasonic image three-dimensional destination object, finished cutting apart then at last to described medical ultrasonic image three-dimensional destination object based on the interpolation method of shape.Compared with prior art, whole cutting procedure has higher segmentation precision and certain automaticity, and has time performance preferably.
The present invention is not only applicable to medical ultrasonic image, is applicable to other medical image sequences, as CT, MRI etc. yet.
The above; it only is a kind of embodiment of the embodiment of the invention; but the protection domain of the embodiment of the invention is not limited thereto; anyly be familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, the protection domain of the embodiment of the invention should be as the criterion with the protection domain of claim.
Claims (12)
1. the dividing method of a medical ultrasonic image three-dimensional destination object is characterized in that, comprising:
To the rearrangement of recombinating of original view data at random, obtain the rule body data set, described original view data at random is the primitive medicine ultrasonic image data that comprise described medical ultrasonic image three-dimensional destination object;
Described rule body data set is carried out cutting apart based on outline line.
2. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 1 is characterized in that, described to the rearrangement of recombinating of original view data at random, the step that obtains the rule body data set comprises:
Original view data at random is carried out the interpolation smoothing processing, make described original view data at random after the processing be filled on the three dimensional network lattice point of the rule that square forms the formation rule volumetric data set.
3. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 1 and 2 is characterized in that, described described rule body data set is carried out comprising based on the step of cutting apart of outline line:
Select key frame in described rule body data centralization;
On described key frame, extract the profile of described medical ultrasonic image three-dimensional destination object;
Extract the profile of described medical ultrasonic image three-dimensional destination object on intermediate frame, the picture frame between the described key frame is an intermediate frame.
4. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 3 is characterized in that, described step at described rule body data centralization selection key frame comprises:
Extract equally spaced two-dimentional parallel image sequence frame in described rule body data centralization, described two-dimentional parallel image sequence frame comprises more than one sequence frame section; Initial and end two frames of described each sequence frame section are key frame.
5. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 4 is characterized in that, the described step of extracting the profile of described medical ultrasonic image three-dimensional destination object on described key frame comprises:
On described key frame, extract the profile of described medical ultrasonic image three-dimensional destination object based on topological self-adaptation deformation T-Snake model.
6. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 5, it is characterized in that the described step of extracting the profile of described medical ultrasonic image three-dimensional destination object based on topological self-adaptation deformation T-Snake model on described key frame comprises:
Triangular grids is carried out in two dimensional image space to described key frame;
Described T-Snake model is initialized as a closed curve, and the intersection point of described T-Snake model and described triangular mesh is the start node of described T-Snake model;
Whether internal force and the external force of judging each node of described T-Snake model all reach balance, if all do not reach balance, then:
Position to each node of described T-Snake model is upgraded, and the normal vector direction of each node along the external force of this node moved, and obtains the T-Snake model after the deformation;
Determine the T-Snake model after the described deformation and the intersection point of described triangular mesh, with its new node of T-Snake model after as described deformation;
Whether internal force and the external force of judging each node of the T-Snake model after the described deformation all reach balance, if all do not reach balance, then repeat above-mentioned steps, till the internal force of each node of T-Snake model and external force all reached equilibrium state, then described profile at the described medical ultrasonic image three-dimensional destination object that extracts on the described key frame was: the matched curve that all nodes of described T-Snake model are formed under the described equilibrium state.
7. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 6 is characterized in that, the step that described T-Snake model is initialized as a closed curve comprises:
Described T-Snake model is initialized as a circle.
8. according to the dividing method of claim 6 or 7 described medical ultrasonic image three-dimensional destination objects, it is characterized in that the closed contour of described T-Snake model for connecting into by N node, described N is a positive integer, and N 〉=2, wherein, the equation of motion of each node is:
If in the above-mentioned equation
Be zero, then the internal force of node and external force reach balance;
Wherein,
Be i (i=0 ..., the N-1) function of position of individual node,
Be the movement velocity of i node, γ i is a ratio of damping;
Internal energy for described T-Snake model comprises tensile force
And bending force
Wherein,
Be the tensile force at i node place,
It is the bending force at i node place; Parameter a is controlling described stretching intensity of force, and parameter b is being controlled described crooked intensity of force, and described internal energy is the flatness constraint condition of described T-Snake model;
Be the external energy of described T-Snake model, this external energy comprises a bulging force
With a external force based on image gradient features
Described bulging force
Direction be the normal vector direction of the external force of node, be the moving direction of the node of described T-Snake model; The external force of described image gradient features
Control model and converging on efficient frontier.
9. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 8 is characterized in that, the tensile force of described each node of T-Snake model and the computing formula of bending force are:
Wherein, d
1Be the distance between i node and i-1 the node, d
2It is the distance between i+1 node and i the node.
10. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 8 is characterized in that, the definite of the normal vector direction of the external force of described each node of T-Snake model specifically comprises:
A node in the described T-Snake model of process reaches two other node with its next-door neighbour, determines a circle, and wherein, described three nodes are not on same straight line;
If the center of circle of described circle is in the border of described T-Snake model, the normal vector direction of the external force of this node is that radial finger along above-mentioned circle is to this node;
If the center of circle of above-mentioned circle is outside the border of described T-Snake model, the normal vector direction of the external force of this node is for radially deviating from this node along above-mentioned circle.
11. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 3 is characterized in that, the described step of extracting the profile of described medical ultrasonic image three-dimensional destination object on intermediate frame comprises:
By on intermediate frame, extract the profile of described medical ultrasonic image three-dimensional destination object based on the interpolation method of shape.
12. the dividing method of medical ultrasonic image three-dimensional destination object according to claim 11 is characterized in that, the described step of extracting the profile of described medical ultrasonic image three-dimensional destination object by the interpolation method based on shape on intermediate frame comprises:
Profile according to the described medical ultrasonic image three-dimensional destination object that extracts on described key frame carries out range conversion to described key frame, obtains key frame distance figure;
Key frame distance figure to two adjacent key frames carries out linear interpolation, obtains the intermediate frame distance map;
Extract the zero point on the described intermediate frame distance map, form the profile of the described medical ultrasonic image three-dimensional destination object on this intermediate frame described each zero point;
Repeat above-mentioned steps, all pass through linear interpolation up to the key frame distance figure of any two adjacent key frames and obtain the intermediate frame distance map, extract the zero point on all intermediate frame distance maps, form the profile of the described medical ultrasonic image three-dimensional destination object on all intermediate frames all zero points.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910005673A CN101807296A (en) | 2009-02-16 | 2009-02-16 | Segmentation method of medical ultrasonic image three-dimensional target object |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN200910005673A CN101807296A (en) | 2009-02-16 | 2009-02-16 | Segmentation method of medical ultrasonic image three-dimensional target object |
Publications (1)
Publication Number | Publication Date |
---|---|
CN101807296A true CN101807296A (en) | 2010-08-18 |
Family
ID=42609079
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN200910005673A Pending CN101807296A (en) | 2009-02-16 | 2009-02-16 | Segmentation method of medical ultrasonic image three-dimensional target object |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN101807296A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102402788A (en) * | 2011-12-22 | 2012-04-04 | 华南理工大学 | Method for segmenting three-dimensional ultrasonic image |
CN102890823A (en) * | 2011-07-19 | 2013-01-23 | 株式会社东芝 | Motion object outline extraction and left ventricular image segmentation method and device |
CN103295234A (en) * | 2013-05-17 | 2013-09-11 | 上海大图医疗科技有限公司 | Medical image segmentation system and medical image segmentation method based on deformation surface models |
CN103824306A (en) * | 2014-03-25 | 2014-05-28 | 武汉大学 | Ultrasonic image segmentation method for dynamics-based statistical shape model |
CN104156975A (en) * | 2013-05-13 | 2014-11-19 | 株式会社东芝 | Medical image analyzing device and method and medical imaging device |
CN104318553A (en) * | 2014-10-15 | 2015-01-28 | 北京理工大学 | Self-adaptive surface deformation model based CT (Computed Tomography) image liver segmentation method |
CN105787978A (en) * | 2016-02-29 | 2016-07-20 | 深圳市医诺智能科技发展有限公司 | Automatic medical image interlayer sketching method, device and system |
CN106651895A (en) * | 2015-11-05 | 2017-05-10 | 沈阳东软医疗系统有限公司 | Method and device for segmenting three-dimensional image |
CN106952264A (en) * | 2017-03-07 | 2017-07-14 | 青岛海信医疗设备股份有限公司 | The cutting method and device of 3 D medical target |
CN108573532A (en) * | 2018-04-16 | 2018-09-25 | 北京市神经外科研究所 | A kind of methods of exhibiting and device, computer storage media of mixed model |
CN109716389A (en) * | 2016-09-05 | 2019-05-03 | 光线搜索实验室公司 | The image processing system and method for interaction outline for 3 D medical data |
-
2009
- 2009-02-16 CN CN200910005673A patent/CN101807296A/en active Pending
Cited By (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102890823A (en) * | 2011-07-19 | 2013-01-23 | 株式会社东芝 | Motion object outline extraction and left ventricular image segmentation method and device |
CN102890823B (en) * | 2011-07-19 | 2016-04-27 | 株式会社东芝 | Motion object outline is extracted and left ventricle image partition method and device |
CN102402788A (en) * | 2011-12-22 | 2012-04-04 | 华南理工大学 | Method for segmenting three-dimensional ultrasonic image |
CN104156975A (en) * | 2013-05-13 | 2014-11-19 | 株式会社东芝 | Medical image analyzing device and method and medical imaging device |
CN103295234A (en) * | 2013-05-17 | 2013-09-11 | 上海大图医疗科技有限公司 | Medical image segmentation system and medical image segmentation method based on deformation surface models |
CN103295234B (en) * | 2013-05-17 | 2016-01-06 | 上海大图医疗科技有限公司 | Based on the medical image segmentation system and method for deformation surface model |
CN103824306B (en) * | 2014-03-25 | 2016-09-14 | 武汉大学 | A kind of ultrasonic image division method based on kinetics statistical shape model |
CN103824306A (en) * | 2014-03-25 | 2014-05-28 | 武汉大学 | Ultrasonic image segmentation method for dynamics-based statistical shape model |
CN104318553B (en) * | 2014-10-15 | 2017-07-14 | 北京理工大学 | CT image liver segmentation methods based on adaptive surface deformation model |
CN104318553A (en) * | 2014-10-15 | 2015-01-28 | 北京理工大学 | Self-adaptive surface deformation model based CT (Computed Tomography) image liver segmentation method |
CN106651895A (en) * | 2015-11-05 | 2017-05-10 | 沈阳东软医疗系统有限公司 | Method and device for segmenting three-dimensional image |
CN106651895B (en) * | 2015-11-05 | 2020-03-17 | 东软医疗系统股份有限公司 | Method and device for segmenting three-dimensional image |
CN105787978A (en) * | 2016-02-29 | 2016-07-20 | 深圳市医诺智能科技发展有限公司 | Automatic medical image interlayer sketching method, device and system |
CN109716389A (en) * | 2016-09-05 | 2019-05-03 | 光线搜索实验室公司 | The image processing system and method for interaction outline for 3 D medical data |
CN109716389B (en) * | 2016-09-05 | 2020-04-14 | 光线搜索实验室公司 | Image processing system and method for interactive contouring of three-dimensional medical data |
CN106952264A (en) * | 2017-03-07 | 2017-07-14 | 青岛海信医疗设备股份有限公司 | The cutting method and device of 3 D medical target |
CN108573532A (en) * | 2018-04-16 | 2018-09-25 | 北京市神经外科研究所 | A kind of methods of exhibiting and device, computer storage media of mixed model |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN101807296A (en) | Segmentation method of medical ultrasonic image three-dimensional target object | |
CN107808377B (en) | The positioning device of lesion in a kind of lobe of the lung | |
CN102068281B (en) | Processing method for space-occupying lesion ultrasonic images | |
US8000941B2 (en) | System and method for surface reconstruction from an unstructured point set | |
CN104240287B (en) | A kind of utilization CT images generate the method and system of coronary artery panorama sketch | |
CN104809730A (en) | Method and device for extracting trachea from chest CT (computed tomography) image | |
CN104835112A (en) | Liver multi-phase CT image fusion method | |
CN102243759B (en) | Three-dimensional lung vessel image segmentation method based on geometric deformation model | |
US8675940B2 (en) | Generation of moving vascular models and blood flow analysis from moving vascular models and phase contrast MRI | |
CN101504767B (en) | Image splitting method based on level set relay | |
CN103353987A (en) | Superpixel segmentation method based on fuzzy theory | |
CN102682449A (en) | Automatic segmentation method based on self-adaptive external force level set for magnetic resonance images (MRIs) of soft tissue and realization method thereof | |
CN101639935A (en) | Digital human serial section image segmentation method based on geometric active contour target tracking | |
CN103310458A (en) | Method for elastically registering medical images by aid of combined convex hull matching and multi-scale classification strategy | |
Le Guyader et al. | Self-repelling snakes for topology-preserving segmentation models | |
CN103914814A (en) | Image fusion method and system for CT coronary image and XA angiography image | |
CN105224764B (en) | Bone modeling and simulation method | |
CN106056596A (en) | Fully-automatic three-dimensional liver segmentation method based on local apriori information and convex optimization | |
CN109886946A (en) | The Weakly supervised classification method of early-stage senile maculopathy based on deep learning | |
Kulp et al. | Using high resolution cardiac CT data to model and visualize patient-specific interactions between trabeculae and blood flow | |
CN106504329A (en) | A kind of gingiva deformity emulation mode of the Mass-spring Model based on tooth long axile | |
CN104318554A (en) | Triangulation optical matching based medical image rigid registration method | |
CN105389821A (en) | Medical image segmentation method based on combination of cloud module and image segmentation | |
CN106548476A (en) | Using medical image statistics pulmonary three-dimensional feature Method On Shape | |
CN102881021B (en) | Based on the aorta petal ultrasonic image division method of probability distribution and continuous max-flow |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C12 | Rejection of a patent application after its publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20100818 |