CN108765385A - A kind of double source CT coronary artery extraction method - Google Patents
A kind of double source CT coronary artery extraction method Download PDFInfo
- Publication number
- CN108765385A CN108765385A CN201810467815.9A CN201810467815A CN108765385A CN 108765385 A CN108765385 A CN 108765385A CN 201810467815 A CN201810467815 A CN 201810467815A CN 108765385 A CN108765385 A CN 108765385A
- Authority
- CN
- China
- Prior art keywords
- coronary artery
- sphere
- angiosomes
- blood vessel
- classes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
Abstract
The present invention relates to a kind of double source CT coronary artery extraction methods, include the following steps:S1. the eigenvalue λ 1 of Hessian matrixes, λ 2 and λ 3 are asked each pixel in double source CT data, new vascular pattern equation is built;Blood vessel structure is effectively enhanced using new vascular pattern;So that angiosomes gray scale is partially bright, background area is partially dark;Gauss modeling is carried out to angiosomes and background area using statistical model;S2. Gaussian mixed models are utilized, enhanced data are divided into 2 classes, background area and angiosomes;S3. in angiosomes, neighborhood relationships are constrained by spherical model growth, grow coronary artery, and the pseudo- branch vessel in the coronary artery that grows is detected and is removed using the branch detection based on hierarchical clustering.
Description
Technical field
The present invention relates to field of medical image processing, more particularly, to a kind of double source CT coronary artery side of automatically extracting
Method.
Background technology
Coronary artery segmentation is the first step of angiocardiopathy image analysis.CT angiographies are current universal inspections
One of mode.CT angiographic data amounts are big, and individual patient difference is big, and cardiovascular structures are complicated.Currently, coronary artery segmentation side
Method, which exists, needs man-machine interactively, and pipeline is susceptible to the problem of pipe branch mistake when reconstructing, directly affect doctor to blood vessel
Analysis.Therefore, most of doctor is still using multiplanar reconstruction (Multi-planner reformation, MPR), curved surface
Rebuild (Curved planner reformation, CPR) and maximum intensity projection (Maximum intensity
Projection, MIP) etc. two dimensional images observe blood vessel, blood vessel situation is analyzed by semi-artificial method to make diagnosis, according to
The subjective experience for relying Yu doctor needs to take a significant amount of time, and is easy delay treatment.
In existing coronary artery extracting method, mainly there are region-growing method, statistical method, movable contour model
Method, center collimation method and multi-scale filtering etc. are simple to judge using the gray feature or morphological feature of blood vessel mostly.Its
In, Frangi is based on multiscale space and proposes three-dimensional blood vessel Enhancement Method, by analyzing hessian matrixes under different scale
Characteristic value judges whether each tissue points belongs to tubular structure [1], but this method is inadequate to the response of vessel borders, and
And it is more sensitive to noise, while enhancing blood vessel, also there is larger enhancing to noise.Tim Jerman improve Frangi
Method, can also successfully be extracted [2] to carrying angiomatous blood vessel.The field of forces GVF are added in the method that Yang improves Frangi
Information can extract entire coronary artery [3], but, tube chamber boundary can not be accurately detected.Zhao Jie et al. in 2016,
Using a kind of blood vessel automatic division method [4] based on multi-scale filtering and Probabilistic Decision-making.Because this method is based on directly on
The multi-scale filtering Enhancement Method of Frangi, therefore, the deficiency of Frangi methods is also embodied in segmentation result.Yang Rong holds high up et al.
Preferably be partitioned into coronary artery from double source CT data, but be split based on each layer of 2-D data, segmentation result into
Row three-dimensional reconstruction blood vessel is not smooth enough [5].Golden shower roc et al. proposes construction spherical model, blood vessel is regarded as in work before
Binary tree structure extracts the center line [6-7] of blood vessel by the rolling of sphere.It, can but when encountering multiple-limb the case where
Pseudo- blood vessel can occur, be taken when traversing blood vessel more.
There is following technological deficiencies when specifically used for said program:
1, simple that coronary artery can not accurately be extracted based on gray feature or morphological feature.Because of coronary blood
Managed network presents extremely complex structure and pathological characteristic, such as the blood vessel size and curvature largely changed, and disease
Reason and degree of impairment under intravascular stent, calcification, aneurysm, luminal stenosis the problems such as.
2, in existing cardiovascular dividing method, certain methods can be very good segmentation bleeding in two dimensional image
Pipe, but model is complicated, speed is slower, is not appropriate for three-dimensional segmentation.Certain methods need man-machine interactively excessive, constrain method
Application.
3, due to coronary artery and its perienchyma's gray scale is more close, motion artifacts and external magnetic field can cause noise dry
It disturbs, the method for Frangi et al. is weaker to vessel borders response, and the method for golden shower roc et al. also deposits the detection of multi-branched blood vessel
In difficulty.
Bibliography
[1]Frangi,A.F.,Niessen,W.J.,Vincken,K.L.,Viergever,M.A.Multiscale
vessel enhancement filtering.Medical Image Computing and Computer-Assisted
Intervention 1496,130-137,http://dx.doi.org/10.1007/BFb0056195(1998).
[2]Tim,J.,Franjo,P.,Bostjan,L.&Ziga,S.Beyond Frangi:an improved
multiscale vesselness filter.International Conference on
Information Science3,http://dx.doi.org/10.1117/12.2081147(2015).
[3]Yang,Y.,Tannenbaum,A.,Giddens,D.&Stillman,A.Automatic segmentation
of coronary arteries using Bayesian driven implicit
surfaces.IEEE International Symposium on Biomedical Imaging5,189-192,
http://dx.doi.org/10.1109/ISBI.2007.356820(2007).
[4] Zhao Jie, Jiang Shizhong, golden shower roc, the research biologies doctor of the emerging double source CTs coronary artery three-dimensional dividing method in Europe Shan
Learn engineering research, 2016,35 (3):197-201.
[5]Ken Cai,Rongqian Yang,Lihua Li,et al.A Semi-Automatic Coronary
Artery Segmentation Framework Using Mechanical Simulation[J].J Med Sys.2015,
39(10):1-7.
[6]Jiang,S.,Xiao,W.,Ou,S.,Huang,Z.&Zhao,J.Three Dimensional Coronary
Artery Centerline Extraction Based on Sphere Model with Branch
Detection.Journal of Computational Information Systems9,9833-9840,http://
dx.doi.org/10.12733/jcis8163(2013).
[7]Huang,Z,Jiang,S.,Liu,Y.,Bao,S.Interactive 3D Liver Vessels
Centerline Extraction based on Moving Sphere Model.International Conference
on Information Science3,2100-2103,http://dx.doi.org/10.1109/
InfoSEEE.2014.6946295(2014).
Invention content
The present invention there are branch detection difficulty and needs manually operated ask for existing heart coronary artery dividing method
Topic, while considering gray feature, morphological feature and the spatial relationship of blood vessel, in conjunction with Statistical Classification method, improve multiple dimensioned
Filtering and spherical model realize that interacting reconnaissance without doctor can realize that three-dimensional coronary is divided automatically.
To realize the above goal of the invention, the technical solution adopted is that:
A kind of double source CT coronary artery extraction method, includes the following steps:
S1. the eigenvalue λ 1 of Hessian matrixes, λ 2 and λ 3 are asked each pixel in double source CT data, is built new
Vascular pattern equation:
WhereinBlood vessel structure is effectively enhanced using new vascular pattern;Make
Angiosomes gray scale it is partially bright, background area is partially dark;Gauss modeling is carried out to angiosomes and background area using statistical model;
S2. Gaussian mixed models are utilized, enhanced data are divided into 2 classes, background area and angiosomes;
S3. in angiosomes, neighborhood relationships are constrained by spherical model growth, grow coronary artery, and utilize base
The pseudo- branch vessel in the coronary artery that grows is detected and is removed in the branch detection of hierarchical clustering.
Preferably, the step S3 is utilized to be detected based on hierarchical clustering and is as follows:
(1) qualified sphere is looked for from root node neighborhood;
(2) it if radius is more than 0, searches using R as all the points of radius, otherwise terminates;
(3) if finding the sphere met, it is R+1, the new sphere of R+2, R+3 to construct radius;Otherwise, R=R-1, and
Recalculate whether radius is more than 0;
(4) it to the ray sent out from the centre of sphere, is clustered with hierarchy clustering method from top to bottom;
(5) if gathering the left child for being arranged that the sphere is present node for 2 classes;If gathered for 3 classes, which is set
Body is the right child of present node;
(6) step (2)-step (5) is repeated, next new node is found, until R<0, terminate.
Compared with prior art, the beneficial effects of the invention are as follows:
(1) prior art is only absorbed in form, gray scale or neighborhood characteristics when extracting blood vessel.The present invention is simultaneously from three sides
Detect blood vessel in face.The form of blood vessel is tubulose, by seeking each pixel the characteristic value of Hessian matrixes, to every individual
The Hessian matrixes of vegetarian refreshments calculate eigenvalue λ 1, λ 2 and λ 3, and characteristic value can be used for distinguishing tubulose, sheet and dots structure.Increase
Image after strong, angiosomes gray scale is partially bright, and background area is partially dark, using statistical model, is carried out to blood vessel class and background classes high
This modeling.In blood vessel class, neighborhood relationships are constrained by spherical model growth, grow coronary artery.In order to avoid pseudo- branch
Blood vessel and the time is saved, establishes the branch detection based on hierarchical clustering.
(2) present invention improves over the tubular structure Enhancement Method that frangi is proposed, the Rb factors is had modified, are replaced with matrix norm
The influence of gray scale unevenness and noise is reduced, and eliminates parameter beta, obtains a new blood vessel equation.
(3) although spherical model can traverse tree construction, the segmentation result of blood vessel is obtained, operation efficiency is also higher.But it is preced with
Shape artery has branch, the branch of mistake that can cause pseudo- blood vessel and expend a large amount of operation times.To solve the above-mentioned problems, originally
Invention proposes the branch detection method based on cluster.Judge the quantity of current location top set by judging the quantity of cluster.
If when the ray that sends out of central point of forecourt gathers for 2 classes, illustrate that current location does not have branch.Ray gathers for 3 classes, then explanation has
Branch.
Description of the drawings
Fig. 1 is the flow chart of method provided by the invention.
Fig. 2 is that ball searches for schematic diagram along blood vessel.
Fig. 3 is the branch detection schematic diagram based on cluster.
Fig. 4 is the spherical model flow chart based on branch detection.
Fig. 5 is arteria coroaria sinistra segmentation result figure.
Fig. 6 is the comparison diagram of blood vessel Enhancement Method.
Fig. 7 is the ROC curve schematic diagram of the present invention and Frangi coronary artery Enhancement Methods.
Fig. 8 is the schematic diagram of coronary artery center line and Calibration.
Specific implementation mode
The attached figures are only used for illustrative purposes and cannot be understood as limitating the patent;
Below in conjunction with drawings and examples, the present invention is further elaborated.
Embodiment 1
The present invention divides coronary artery from three vascular morphology, gray feature and neighborhood relationships aspects simultaneously.Blood
The form of pipe is tubulose, and by seeking each pixel the characteristic value of Hessian matrixes, the Rb for improving frangi models is calculated
Son establishes new vascular pattern to enhance blood vessel structure.Enhanced image, angiosomes gray scale is partially bright, and background area is partially dark,
Using statistical model, Gauss modeling is carried out to blood vessel class and background classes.In blood vessel class, neighborhood is constrained by spherical model growth
Relationship grows coronary artery.In order to avoid pseudo- branch vessel and the time is saved, establishes branch's inspection based on hierarchical clustering
It surveys.The overview flow chart of concrete methods of realizing of the present invention is as shown in Figure 1.
Present invention improves over the blood vessels of Frangi to enhance model.Since the method for the Frangi enhancing blood vessels proposed is to noise
Compare sensitive, while noise can be enhanced, therefore, it is necessary to do some improvement, to inhibit ambient noise.The present invention has modified frangi
The Rb factors in method are replaced with matrix norm to reduce the influence of gray scale unevenness and noise, and it is new to eliminate parameter beta mono-
Blood vessel equation be defined as follows:
Wherein
Assuming that each class loading meets same Gaussian Profile in image, then in image all classification can be regarded as it is multiple high
The distribution of the sum of this distribution, entire volume data is formed by the gaussian probability linear combination of each tissue class, as shown in formula (2),
wkFor the weight shared by every class loading.
To enhanced image, corresponding low gray area is background, and corresponding high gray scale is blood area.Therefore, it can incite somebody to action
Data are divided into 2 classes.Total data probability distributions are formed by stacking by 2 histioid Gaussian Profiles, as shown in formula (3).
Parameter wk, uk, σkEstimate by greatest hope method.
E steps calculate the cluster probability of each example using distributed constant, and as shown in formula (4), t represents iterative steps.
M steps reevaluate distributed constant with maximal possibility estimation, obtain:
All parameters of gauss hybrid models are obtained by EM methods, then, by the way which belongs to the differentiation of each voxel
One class loading, criterion belong to kth class when being k ≠ j.
When extracting coronary artery center line, since some designated position coronarius, according to preassigned item
Part, the spheroid that one size of imagination can be changed move ahead along coronary flow direction, and according to blood vessel at its location
The size of internal diameter constantly adjusts sphere size so that current spheroid is the maximum inscribed spheroid of radius at the position, when ellipse
When sphere encounters coronary artery bifurcation in moving process, current spheroid is split into two spheroids, respectively along coronal
Each branch of artery continues according to said method to move ahead, and method stops when searching for the spheroid less than the condition that meets, and finally will
The centre of sphere of the obtained spheroid on each position connects, this connecting line just constitutes entire center coronarius
Line, as shown in Figure 2.
With the progress of search, spheroid, which needs to move ahead along blood direction in blood vessel, could complete entire vessel centerline
Extraction.As seen from Figure 2, after spheroid is inscribed in the maximum of current location, the centre of sphere of next spheroid can be current
It is chosen on the half ellipsoidal surface of the inscribed spheroid of maximum, therefore the moving step length of spheroid is exactly the current maximum inscribed spheroid centre of sphere
The distance of each point on to half ellipsoidal surface.For each point on half ellipsoidal surface, put using this as next inscribed spheroid
The centre of sphere calculates and spheroid is inscribed by the maximum of the centre of sphere of each point according to the threshold condition of formula (8), is finally corresponded to from each point
Maximum be inscribed ellipsoid in select the maximum spheroid of volume to get to the maximum of next step be inscribed spheroid.
Wherein, p (x, y, z) is the coordinate of any point, and g (x, y, z) is the gray scale of point p (x, y, z).By many experiments,
The empirical value h of coronary artery is between 0.9-0.95.
Although spherical model can traverse tree construction, the segmentation result of blood vessel is obtained, operation efficiency is also higher.But it is coronal
Artery has branch, the branch of mistake that can cause pseudo- blood vessel and expend a large amount of operation times.When balls tumble to vessel branch
When, how to determine numbers of branches, and it is a problem to be searched for according to correct branch.
To solve the above-mentioned problems, the present invention proposes the branch detection method based on cluster.As shown in figure 3, from ball
Heart point sends out ray, and the radius of ball is R, is tested respectively with R+1, R+2 and R+3 construction sphere.Reach outermost layer sphere
Ray gathers for several classes, judges the quantity of current location top set by judging the quantity of cluster.If when the central point of forecourt
The ray sent out gathers for 2 classes, illustrates that current location does not have branch.As in Fig. 3, ray 1 and ray 2 gather for 3 classes, then explanation has point
Branch.
General clustering method needs selected k values, i.e., determination is divided into several classes before classification, & apos.But since blood vessel structure is multiple
Miscellaneous, individual difference is big, the presence of patient's blood vessel deformity, is difficult to determine to be divided into several classes, and hierarchy clustering method can be with before cluster
This is solved the problems, such as well.
Assuming that the ray for having N items to be clustered, for top-down hierarchical clustering (agglomerative hierarchical
Clustering for), basic step is:
Step 1, initialization.Every ray sample is classified as one kind, calculates the distance between each two class, that is, sample
Similarity between sample;
The distance between step 2, each class of calculating, are arranged a threshold value, when the distance of two nearest classes is more than this
Threshold value, then it is assumed that iteration can terminate.
Step 3 finds out two nearest classes, they are classified as one kind;
Step 4 recalculates newly-generated similar the distance between all kinds of, and jumps to step 2.
The Euclidean distance between ray and each point of outer surface of spheroid intersection is calculated, using SingleLinkage methods
Calculate the distance between two classes, the distances of two samples for exactly taking distance in two classes nearest as the two gather away from
From.Single linkage methods are as shown in formula 9.
Dmin=min | p1-p2| formula (9)
Wherein, p1, p2 are two points in inhomogeneity.
As the distance between fruit is more than by following formula (10) calculated distance, then cluster operation termination.
Wherein, α is the angle of ray 1 and ray 2, as shown in Figure 3.
The flow chart for the 3D vessel branch detection methods that the present invention designs is as shown in Figure 4.Method is as follows:
(1) qualified sphere is looked for from root node neighborhood;
(2) it if radius is more than 0, searches using R as all the points of radius, otherwise terminates;
(3) if finding the sphere met, it is R+1, the new sphere of R+2, R+3 to construct radius;Otherwise, R=R-1, and
Recalculate whether radius is more than 0;
(4) it to the ray sent out from the centre of sphere, is clustered with hierarchy clustering method from top to bottom;
(5) if gathering the left child for being arranged that the sphere is present node for 2 classes;If gathered for 3 classes, which is set
Body is the right child of present node;
(6) step (2)-step (5) is repeated, next new node is found, until R<0, method terminates.
Present invention combination statistics, multiscale space and the spherical model method based on gray scale, consider gray scale coronarius
Feature, morphological feature and spatial relation can accurately be partitioned into coronary artery automatically.As shown in figure 5, the left hat being partitioned into
Shape artery reconstruction effect is preferable, avoids blood vessel leakage problem, and without pseudo- blood vessel, arteriarctia is high-visible.
Blood vessel proposed by the present invention enhances new method, carries out Hessian operations to each pixel of volume data first, recycles
Formula (4) detects blood vessel, obtains Fig. 6 (b), it can be seen that tubular structure is enhanced.Fig. 6 (c) is the blood proposed with Frangi
The image that pipe Enhancement Method obtains.By the Comparative result to Fig. 6 (b) and Fig. 6 (c), (b) in without apparent hole and disconnected
It splits, enhancing effect is better than (c).
It is commented with Receiver operating curve (Receiver Operating Characteristic Curve, ROC)
Valence both methods, obtains Fig. 7.Blue is the evaluation result of the present invention, and red is the evaluation result of Frangi, it can be seen that this
Improved effect is invented, area under a curve (Area Under the Curve, AUC) becomes larger, and evaluation effect is better than Frangi
Method.
Difficult, branch detection method of the present invention proposition based on cluster is extracted to coronary artery multi-branched for spherical model.
The coronary artery extracted in Fig. 8 obviously has narrow, and stenosis is only that 1~2 pixel is wide, but method proposed by the present invention still may be used
Correctly to divide, it is not broken.
Obviously, the above embodiment of the present invention be only to clearly illustrate example of the present invention, and not be pair
The restriction of embodiments of the present invention.For those of ordinary skill in the art, may be used also on the basis of the above description
To make other variations or changes in different ways.There is no necessity and possibility to exhaust all the enbodiments.It is all this
All any modification, equivalent and improvement etc., should be included in the claims in the present invention made by within the spirit and principle of invention
Protection domain within.
Claims (2)
1. a kind of double source CT coronary artery extraction method, it is characterised in that:Include the following steps:
S1. the eigenvalue λ 1 of Hessian matrixes, λ 2 and λ 3 are asked each pixel in double source CT data, new blood vessel is built
Model equation:
WhereinBlood vessel structure is effectively enhanced using new vascular pattern;Make blood
Pipe area grayscale is partially bright, and background area is partially dark;Gauss modeling is carried out to angiosomes and background area using statistical model;
S2. Gaussian mixed models are utilized, enhanced data are divided into 2 classes, background area and angiosomes;
S3. in angiosomes, neighborhood relationships are constrained by spherical model growth, grow coronary artery, and using based on layer
The branch detection of secondary cluster is detected and removes to the pseudo- branch vessel in the coronary artery that grows.
2. double source CT coronary artery extraction method according to claim 1, it is characterised in that:The step S3 is utilized
It is detected and is as follows based on hierarchical clustering:
(1) qualified sphere is looked for from root node neighborhood;
(2) it if radius is more than 0, searches using R as all the points of radius, otherwise terminates;
(3) if finding the sphere met, it is R+1, the new sphere of R+2, R+3 to construct radius;Otherwise, R=R-1, and again
Calculate whether radius is more than 0;
(4) it to the ray sent out from the centre of sphere, is clustered with hierarchy clustering method from top to bottom;
(5) if gathering the left child for being arranged that the sphere is present node for 2 classes;If gathered for 3 classes, the sphere, which is arranged, is
The right child of present node;
(6) step (2)-step (5) is repeated, next new node is found, until R<0, terminate.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810467815.9A CN108765385B (en) | 2018-05-14 | 2018-05-14 | Double-source CT coronary artery automatic extraction method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810467815.9A CN108765385B (en) | 2018-05-14 | 2018-05-14 | Double-source CT coronary artery automatic extraction method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108765385A true CN108765385A (en) | 2018-11-06 |
CN108765385B CN108765385B (en) | 2021-12-07 |
Family
ID=64008138
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810467815.9A Active CN108765385B (en) | 2018-05-14 | 2018-05-14 | Double-source CT coronary artery automatic extraction method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108765385B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109472807A (en) * | 2018-11-30 | 2019-03-15 | 北京师范大学 | Vascular pattern extracting method based on deep neural network |
CN110458848A (en) * | 2019-07-05 | 2019-11-15 | 心医国际数字医疗系统(大连)有限公司 | Judge the method that coronary branches and mask data are converted to center line point set in the coronary artery segmentation of image |
CN111612743A (en) * | 2020-04-24 | 2020-09-01 | 杭州电子科技大学 | Coronary artery central line extraction method based on CT image |
CN111932497A (en) * | 2020-06-30 | 2020-11-13 | 数坤(北京)网络科技有限公司 | Coronary artery identification method and device |
CN112308846A (en) * | 2020-11-04 | 2021-02-02 | 赛诺威盛科技(北京)有限公司 | Blood vessel segmentation method and device and electronic equipment |
CN113129301A (en) * | 2021-05-11 | 2021-07-16 | 首都医科大学附属北京天坛医院 | Prediction method, device and equipment for intracranial aneurysm surgical planning |
WO2021249439A1 (en) * | 2020-06-09 | 2021-12-16 | Shanghai United Imaging Healthcare Co., Ltd. | Systems and methods for image processing |
CN117547353A (en) * | 2024-01-12 | 2024-02-13 | 中科璀璨机器人(成都)有限公司 | Tumor accurate positioning and robot puncturing method and system for dual-source CT imaging |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN206183289U (en) * | 2016-06-02 | 2017-05-24 | 黄毅 | Two source CT scanning system |
US20170294034A1 (en) * | 2016-04-11 | 2017-10-12 | Toshiba Medical Systems Corporation | Apparatus and method of iterative image reconstruction using regularization-parameter control |
-
2018
- 2018-05-14 CN CN201810467815.9A patent/CN108765385B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170294034A1 (en) * | 2016-04-11 | 2017-10-12 | Toshiba Medical Systems Corporation | Apparatus and method of iterative image reconstruction using regularization-parameter control |
CN206183289U (en) * | 2016-06-02 | 2017-05-24 | 黄毅 | Two source CT scanning system |
Non-Patent Citations (2)
Title |
---|
蒋世忠: "基于球模型的三维冠状动脉中心线抽取方法研究", 《中国博士学位论文全文数据库医药卫生科技辑》 * |
赵洁等: "双源CT冠状动脉三维分割算法的研究", 《生物医学工程研究》 * |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109472807A (en) * | 2018-11-30 | 2019-03-15 | 北京师范大学 | Vascular pattern extracting method based on deep neural network |
CN109472807B (en) * | 2018-11-30 | 2021-11-26 | 北京师范大学 | Blood vessel model extraction method based on deep neural network |
CN113240684A (en) * | 2019-07-05 | 2021-08-10 | 心医国际数字医疗系统(大连)有限公司 | Method for converting extracted mask data containing coronary artery central line into central line point set |
CN110458848A (en) * | 2019-07-05 | 2019-11-15 | 心医国际数字医疗系统(大连)有限公司 | Judge the method that coronary branches and mask data are converted to center line point set in the coronary artery segmentation of image |
CN110458848B (en) * | 2019-07-05 | 2021-11-26 | 心医国际数字医疗系统(大连)有限公司 | Method for judging coronary artery branch in image coronary artery segmentation |
CN111612743A (en) * | 2020-04-24 | 2020-09-01 | 杭州电子科技大学 | Coronary artery central line extraction method based on CT image |
WO2021249439A1 (en) * | 2020-06-09 | 2021-12-16 | Shanghai United Imaging Healthcare Co., Ltd. | Systems and methods for image processing |
CN111932497A (en) * | 2020-06-30 | 2020-11-13 | 数坤(北京)网络科技有限公司 | Coronary artery identification method and device |
CN112308846B (en) * | 2020-11-04 | 2021-07-13 | 赛诺威盛科技(北京)股份有限公司 | Blood vessel segmentation method and device and electronic equipment |
CN112308846A (en) * | 2020-11-04 | 2021-02-02 | 赛诺威盛科技(北京)有限公司 | Blood vessel segmentation method and device and electronic equipment |
CN113129301A (en) * | 2021-05-11 | 2021-07-16 | 首都医科大学附属北京天坛医院 | Prediction method, device and equipment for intracranial aneurysm surgical planning |
CN117547353A (en) * | 2024-01-12 | 2024-02-13 | 中科璀璨机器人(成都)有限公司 | Tumor accurate positioning and robot puncturing method and system for dual-source CT imaging |
CN117547353B (en) * | 2024-01-12 | 2024-03-19 | 中科璀璨机器人(成都)有限公司 | Tumor positioning method and system for dual-source CT imaging |
Also Published As
Publication number | Publication date |
---|---|
CN108765385B (en) | 2021-12-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108765385A (en) | A kind of double source CT coronary artery extraction method | |
CN109035255B (en) | Method for segmenting aorta with interlayer in CT image based on convolutional neural network | |
CN105741251B (en) | A kind of blood vessel segmentation method of Hepatic CT A sequence images | |
CN106204587B (en) | Multiple organ dividing method based on depth convolutional neural networks and region-competitive model | |
CN107563983A (en) | Image processing method and medical imaging devices | |
US9996918B2 (en) | Method for distinguishing pulmonary artery and pulmonary vein, and method for quantifying blood vessels using same | |
CN105719295B (en) | A kind of intracranial hemorrhage region segmentation method and system based on three-dimensional super voxel | |
Zhuge et al. | An abdominal aortic aneurysm segmentation method: Level set with region and statistical information | |
CN107909581A (en) | Lobe of the lung section dividing method, device, system, storage medium and the equipment of CT images | |
CN108961261B (en) | Optic disk region OCT image hierarchy segmentation method based on space continuity constraint | |
WO2008097520A2 (en) | System and method for computer aided detection of pulmonary embolism in tobogganing in ct angiography | |
Wang et al. | A segmentation method of coronary angiograms based on multi-scale filtering and region-growing | |
CN106056596B (en) | Full-automatic three-dimensional liver segmentation method based on local prior information and convex optimization | |
CN109166095A (en) | A kind of ophthalmoscopic image cup disk dividing method based on generation confrontation mechanism | |
Gerard et al. | Pulmonary lobe segmentation using a sequence of convolutional neural networks for marginal learning | |
Cheng et al. | Automated delineation of calcified vessels in mammography by tracking with uncertainty and graphical linking techniques | |
Tan et al. | Analysis of segmentation of lung parenchyma based on deep learning methods | |
CN111784701A (en) | Ultrasonic image segmentation method and system combining boundary feature enhancement and multi-scale information | |
CN111047607A (en) | Method for automatically segmenting coronary artery | |
CN110458848A (en) | Judge the method that coronary branches and mask data are converted to center line point set in the coronary artery segmentation of image | |
CN109727227A (en) | A kind of diagnosis of thyroid illness method based on SPECT image | |
Zhang et al. | DDNet: a novel network for cerebral artery segmentation from MRA IMAGES | |
Ma et al. | A novel automatic coronary artery segmentation method based on region growing with annular and spherical sector partition | |
Wang et al. | An automated method for segmentation of coronary arteries in coronary ct imaging | |
Ong et al. | Automatic Glaucoma Detection from Stereo Fundus Images |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |