CN104751438B - For the method and apparatus alternatively to medical threedimensional images registration - Google Patents
For the method and apparatus alternatively to medical threedimensional images registration Download PDFInfo
- Publication number
- CN104751438B CN104751438B CN201310746488.8A CN201310746488A CN104751438B CN 104751438 B CN104751438 B CN 104751438B CN 201310746488 A CN201310746488 A CN 201310746488A CN 104751438 B CN104751438 B CN 104751438B
- Authority
- CN
- China
- Prior art keywords
- reference picture
- threedimensional images
- medical threedimensional
- mutual information
- medical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
Disclose a kind of method and apparatus for being alternatively registrated to medical threedimensional images.This method comprises: manually adjust medical threedimensional images with mutual information of the reference picture on multiple vertical direction in spaces based on medical threedimensional images, and at predetermined intervals and/or adjusts number in adjustment process and recalculate the mutual information；Medical threedimensional images are manually adjusted based on the mutual information recalculated, to obtain desired initial registration position；And based on the initial registration position by medical threedimensional images and reference picture automatically accuracy registration, wherein the accuracy registration is carried out using association relationship on multiple vertical direction in spaces of medical threedimensional images and reference picture.
Description
Technical field
The present invention relates to Medical Image Processings, more specifically to the registration of medical threedimensional images.
Background technique
During clinical application or medical research, it is often necessary to be radiated to 3-D image, the three-dimensional of such as skeletal tissue
Image compares.Such three-dimensional radiation image is, for example, spiral CT three-dimensional rebuilding image.It is clinically frequently necessary to patient
The pretherapy and post-treatment image of illness bone compare, assess therapeutic effect.In medicine case study, it is often necessary to by patient
The image of the image of block bone bone identical as normal person compares, it is expected that therefrom finding some pathology rules.
Since the three-dimensional radiation image of skeletal tissue has the following characteristics that outer layer compact bone substance is rounder and more smooth, internal bone trabecula
It in irregular network-like, therefore can not find effective characteristic point in this structure, be difficult to be carried out with the method for feature point alignment
Registration.And since 3 d image data amount is very big, it is also not suitable for being registrated with the method for transform domain.One under such case
As using surface model registration or the registration based on grayscale information.Surface model registration needs first to be partitioned into target bone, then
Surface model is extracted, is registrated using surface shape.Due to absolutely not considering skeletal tissue in surface model registration process
Internal structure, therefore the degree of registration of internal trabecular bone structure may be poor.Registration based on grayscale information considers full figure
Grayscale information, but operand is big, and calculating speed is slow.
Summary of the invention
Therefore there are the needs of the method for registering rapidly and efficiently for medical threedimensional images, in particular for skeletal tissue
Medical threedimensional images, the bone trabecula tissue of bones all can accurately be aligned by this method.
The needs are met by a kind of method for being alternatively registrated to medical threedimensional images.This method
Comprising steps of manually adjusting doctor with mutual information of the reference picture on multiple vertical direction in spaces based on medical threedimensional images
Learn 3-D image, and in adjustment process at predetermined intervals and/or adjust number recalculate the mutual information；Base
Medical threedimensional images are manually adjusted in the mutual information recalculated, to obtain desired initial registration position；And based on institute
Initial registration position is stated by medical threedimensional images and reference picture automatically accuracy registration, wherein the accuracy registration utilizes medicine
The association relationship of 3-D image and reference picture on multiple vertical direction in spaces carries out.
The needs are also met by a kind of equipment for being alternatively registrated to medical threedimensional images, the equipment packet
It includes: display device, for showing reference picture and medical threedimensional images to be registered；Apparatus for adjusting position, for by operator
Medical 3 D figure is manually adjusted with mutual information of the reference picture on multiple vertical direction in spaces based on medical threedimensional images
Picture；Mutual information computing device, at predetermined intervals and/or being adjusted described in number constantly calculates in adjustment process
Mutual information；And accuracy registration device, for being existed based on the initial registration position using medical threedimensional images and reference picture
Association relationship on multiple vertical direction in spaces is by medical threedimensional images and reference picture automatically accuracy registration.
This method carries out pre-align to medical threedimensional images using the mode of interaction first.After the completion of pre-align, it will obtain
Result as initial registration position, then carry out automatic and accurate registration, calculate the time it is possible thereby to shorten, and obtain good
Registration result.Come during accuracy registration using the association relationship of multiple, preferably 3 vertical direction in spaces sectioning images
It is registrated, all voxels of entire 3-D image is calculated therefore without as traditional method for registering based on mutual information
Mutual information, to improve calculating speed.And for the 3-D image of skeletal tissue, moreover it is possible to make bones
Bone trabecula tissue all match very neat.Especially during interactive pre-align, user can manually adjust medicine three
It ties up the position of image and adjusts the position according to the feedback of image display, by medical threedimensional images pre-align, and with
The preferable registration position obtained during pre-align is registrated as initial registration position for next automatic and accurate.It is this
Interactive mode, which is adjusted, can exclude most of local optimum position in the automatic and accurate registration stage later, improve search and calculate
The constringency performance of method, to save the calculating time.
Detailed description of the invention
Fig. 1 shows the flow chart of the embodiment of the method for medical threedimensional images to be interacted with formula registration.
Fig. 2 shows the exemplary process diagrams for the step of initial registration position is obtained in method shown in Fig. 1.
Fig. 3 a schematically shows the adjusting of the registration position on a direction in space.
Fig. 3 b schematically shows the positive face depth map in step shown in Fig. 2.
Fig. 4 shows the exemplary process diagram for the step of accuracy registration is carried out in method shown in Fig. 1.
Fig. 5 shows the dimensional rendered views of reference picture and image subject to registration in pre-align step.
Fig. 6 shows the block diagram of the embodiment of the equipment for interactive registration.
Specific embodiment
Fig. 1 shows the process of the method 100 for medical threedimensional images to be interacted with formula registration.First in step 101
In, floating image and reference picture are shown on the image display of such as computer screen.Then based on more
The association relationship of the sectioning image of a vertical direction is allowed to and reference picture pre-align by manually adjusting floating image.The hand
Dynamic adjusting can for example make image translation and/or rotation to carry out by user's actuated position regulating device, which adjusts dress
Setting is a kind of input equipment, the new position, such as mouse, keyboard, control stick etc. for inputting floating image.The position refers to three
Tie up the pose of image in three dimensions.The floating image, that is, medical threedimensional images to be registered.Multiple vertical direction are preferred
It is common 3 mutually perpendicular direction in space-transverse directions, coronal direction and sagittal direction in medical image.In the step
In, user can be interacted by user interface with image display, per (such as 100ms) and/or every tune at regular intervals
Section once just calculate a floating image and the cross section of reference picture, coronal-plane, three vertical direction sections of sagittal plane it is mutual
The value of information, and using obtained association relationship as the degree of registration metric of floating image and reference picture.The degree of registration
Metric can be used as the reference manually adjusted again in turn.
Then in a step 102, when association relationship reaches satisfactory or scheduled size and/or user passes through naked eyes
When observation determines that the degree of registration of floating image and reference picture is preferable, corresponding registration position is determined as floating image and ginseng
The initial registration position of image is examined, it then can be using the initial registration position as the initial position of next step accuracy registration.On
The position that floating image can be significantly adjusted by the interactive adjusting of user and image display is stated, especially floating
In the biggish situation of the position difference of motion video and reference picture, to eliminated in next automatic and accurate registration well
Most of local optimum position, the convergence of the searching algorithm in accuracy registration is thus improved, when calculating is greatly saved
Between.
Finally in step 103, since the initial registration position, the mutual trust of the sectioning image based on 3 vertical direction
Floating image and reference picture are carried out accuracy registration by the method for iteration by breath value automatically.In this step, by floating image
With the cross section of reference picture, coronal-plane, three directions of sagittal plane sectioning image association relationship as objective function, use
The Parzen window function export search gradient of limited sampling point, iteratively seeks optimal solution with Powell search method.The search is also
It will be described in further detail below.Automatic and accurate registration is the fine tuning and supplement to pre-align manual in step 101.Due in iteration
Each step in, it is only necessary to calculate cross section, coronal-plane, three vertical direction of sagittal plane sectioning image association relationship, with biography
The method for registering based on association relationship of system is compared and simplifies operation, therefore will soon obtain registration result.
Fig. 2 shows the exemplary process diagrams of step 101 in the method 100.
By in the step 101 of floating image and reference picture pre-align, it is first carried out and obtains floating image to be registered
With the step 1011 of the volume textures coordinate of reference picture.Floating image can be manually adjusted thus, such as it is made to translate or rotate.
This, which is manually adjusted, to carry out by the apparatus for adjusting position of such as mouse.The adjusting is shown in a schematic manner in fig. 3 a.
Left figure shows the operation chart on a section of reference picture and floating image, and right figure shows operation instruction icon.In
It is to generate volume data position transformation matrix mat,
Wherein (cx, cy, cz) it is rotation center, (x, y, z) is unit rotation axis vector, and s=sin α, c=cos α, α are to work as
Preceding rotation angle；(Tx, Ty, Tz) it is translation vector.
To realize inquiry of the voxel value in volume textures, need the slice plane coordinate transformation under world coordinate system to be line
Manage the volume textures coordinate under coordinate system.Transition matrix w2t are as follows:
Wherein,
Wherein (ox, oy, oz) be volume data starting point coordinate, (sx, sy, sz) be volume data Pixel Dimensions, (dx, dy, dz)
For the dimension of volume data.
After by adjusting the new position for obtaining floating image, reference picture is obtained in step 1012 and is hung down at 3
The upward axial direction of histogram cuts plane, then by the corresponding axial positive face for cutting plane of rendering, obtains in step 1013
Positive face depth map.The color value of each pixel on positive face depth map represents the corresponding axial depth cut in plane
Value.Fig. 3 b shows the axial depth map for cutting face in a schematic manner.As shown in Figure 3b, current axial direction cuts plane and view
Line direction is vertical.In the process, it retrieves and accordingly axially cuts position of the every bit in world coordinate system in plane, by it
It is converted into texture coordinate:
pt=w2tph,
phPosition coordinates (four-vector), p are cut for what is retrievedtFor the volume textures coordinate under texture coordinate system.
Then by positive face depth map obtained and reference picture and floating image, joined in step 1014
Image and floating image are examined in the texture tile of 3 vertical direction.Here, being obtained for example, by using the method for volume drawing light projection
The sectioning image (being herein texture tile image) of reference picture and floating image in 3 vertical direction.Then for example, by using more
Post-processing object (Multiple Render Targets) technology generates the MI (Mutual of these slice image datas
Information it) calculates copy 1 and copy 2 is drawn in fusion.So-called copy refers to storage copy of the sectioning image in video memory.
During obtaining sectioning image, carried out simultaneously using CUDA (Compute Unified Device Architecture)
Row accelerates.Therefore, because the drafting for being related to cache object and cache object be to the different purposes of two kinds of the mapping of CUDA address space,
To guarantee safe operation, there are two parts of copies in video memory.
Reference picture and floating image are obtained after the sectioning image of 3 vertical direction, cuts these in step 1015
The copy 1 of picture is sliced for calculating association relationship.Since copy 1 is stored in video memory, data are avoided from video memory to interior
The copy deposited, improves calculating speed.Cache object (bufferobject) in addition can be mapped to the address space of CUDA,
To avoid the intervention of CPU.Copy 2 is sliced to the fusion drafting for being used for reference picture and floating image simultaneously.The fusion is drawn
It can be shown to user, so that user checks initial registration degree.
The calculating of association relationship is briefly described below.
The association relationship MI (A, B) of two images A and B is defined as:
Wherein PABRefer to joint probability density, PA、PBIt is the marginal probability density of image A and B.
PABAnd PA、PBIt can be obtained by statistics grey level histogram.
In one example, following formula is used
Joint image is converted by image A and B, wherein x is the space coordinate [x y z] of joint image,It indicates to floating
The break-in operation of points a.The codomain of this joint image I (x) is [0, B1×B2], wherein statistics with histogram group number is B.With parallel
Joint histogram is projected along two axial directions, obtains edge histogram by counting statistics joint histogram.
In this example, the parallel computation of joint histogram is in two stages: in the first phase, each parallel thread block
The histogram of data handled by it will be calculated.Since per thread block is all independent from each other when executing this operation,
These histograms can be calculated in shared drive (Shared Memory).Shared drive resides in physics GPU (Graphics
Processing Unit) on, without being resident in the Installed System Memory except GPU.Therefore, prolonging when accessing shared drive
It late will be well below the delay in access ordinary buffer area.By using shared drive, avoid every time sending out write operation from chip
It is sent to DRAM (Dynamic Random Access Memory).The interim histogram of per thread block is closed in second stage
And into global buffer, color histogram is obtained.By the way that the calculating of histogram is divided into two stages, can greatly reduce
Using there is a situation where compete between thousands of threads when global memory.
At regular intervals calculate cross section, coronal-plane, three vertical direction of sagittal plane sectioning image association relationship,
The degree of registration metric obtained as current operation.User observes the floating image and reference picture that synchronous fusion is drawn, and
And under the auxiliary of the association relationship of acquisition, a preferable initial registration pose is determined.As shown in step 1016, for example, if
Association relationship reaches satisfactory or scheduled size, and the corresponding fusion drawing three-dimensional image of user's observation confirms currently
Operation is so that floating image is preferably aligned with reference picture, then true by the registration position of the corresponding floating image of the association relationship
It is set to initial registration position.If association relationship is unsatisfactory or user observes floating image not with reference picture preferably
Alignment, then continue to manually adjust floating image.
It should be noted that the data that copy 1 is sliced in this example need before transmission function carries out sort operation
Interception.Copy 2 is carrying out fusion drafting after transmission function classification coloring.Transmission function (Transfer Function) is
The sampled point of 3 d data field is mapped as optical parameter by the common technology in volume drawing.
Fig. 4 shows the flow chart that the step 103 of automatic and accurate registration is carried out in method shown in Fig. 1.
In step 103, since the initial registration position obtained during pre-align, automatic precision is carried out to floating image
Really registration.It is not necessarily to the participation of user in the process.
In automatic and accurate registration, extraction reference picture and floating image are in cross section, coronal-plane, sagittal plane three first
The pixel of the sectioning image of a vertical direction section.The pixel of the sectioning image of three vertical direction sections of reference picture can be straight
It connects for calculating mutual information, and the respective slice image pixel of three vertical direction sections of floating image is linearly inserted in progress three
For calculating mutual information after value.It was described during the calculating of association relationship pre-align shown in Fig. 2.
It is used using association relationship as objective function using the Parzen window function export search gradient of limited sampling point
Powell searching method seeks optimal solution.Powell searching method is iterative search method.In the embodiment shown in fig. 4, exist
In each round iteration, first successively searched for along initial 3 vertical direction, obtain one it is most better, then along epicycle iteration
The initial point line direction most better with this scans for, acquire this wheel iteration as a result, being taken again with the last direction of search
For one of preceding 3 directions, start the iteration of next round.
All determine whether the variation of mutual information tends towards stability after every wheel iteration.If the variation of mutual information still compared with
Greatly, then transformation parameter, i.e. the respective slice image pixel of the three of floating image vertical direction section carry out new pixel rigid
Property variation, and carry out Tri linear interpolation again, recalculate association relationship.Until reaching a more stable association relationship,
The stable association relationship illustrates that floating image and being registrated for reference picture are preferable, and so far algorithm terminates.
Fig. 5 shows the dimensional rendered views of reference picture and floating image in pre-align step 101.The figure is respectively on a left side
Top shows the fusion of reference picture and floating image in adjustment process and draws, and shows reference picture and floating figure in upper right side
The cross-sectional view strength of picture shows the coronal-plane view of reference picture and floating image in lower left, shows in lower right with reference to figure
The sagittal plane view of picture and floating image.
Scheme disclosed herein is divided into two stages: i.e. interactive to adjust registration stage and automatic and accurate registration stage.It is first
It is first adjusted in the registration stage in interactive mode, the position of medical threedimensional images is manually adjusted by user and according to image display
Feedback adjust the position, medical threedimensional images and reference picture are subjected to pre-align, with obtained during pre-align compared with
Good registration position is registrated the stage for next automatic and accurate as initial registration position.The stage is registrated in automatic and accurate
In, using association relationship as objective function, optimal solution is iteratively searched using the searching algorithm of such as Powell.Interactive mode, which is adjusted, matches
The quasi- stage can exclude the big portion in the automatic and accurate registration stage later by manually adjusting the position of medical threedimensional images
Branch office's portion's optimal location, improves the constringency performance of searching algorithm, to save the calculating time.In addition, either pre-align
Process or later accuracy registration process are all only to three vertical direction in space-transverse directions, coronal direction and sagittal side
Upward sectioning image calculates mutual information, reduces calculation amount, simplifies operation, to improve search efficiency.The scheme
Alignment and registration especially suitable for skeletal tissue's 3-D image.
The method can be implemented by computer software or hardware or combinations thereof.
Fig. 6 shows the embodiment of the equipment 600 for interactive registration in the form of structural block diagram.The equipment 600 is main
Including apparatus for adjusting position 610, mutual information computing device 630, accuracy registration device 650 and display device 670.
Apparatus for adjusting position 610 is a kind of input unit, such as mouse, keyboard, control stick etc., for being adjusted manually by user
Save the pose of medical threedimensional images to be registered in space.This is manually adjusted can be based on the sectioning image of multiple vertical direction
Association relationship carry out, be also based on the observation of user additionally to carry out.Display device 670 shows to be registered simultaneously
Medical threedimensional images and reference picture.User can observe medical threedimensional images (i.e. floating figure to be registered in display device 670
Picture) with the difference of reference picture, pass through the position that apparatus for adjusting position 610 adjusts floating image.In the process, mutual information meter
It calculates device 630 at predetermined intervals and/or scheduled adjusting number (such as primary every 100ms and/or every adjusting) is counted
Calculate the association relationship of the sectioning image of multiple vertical direction.The association relationship is shown to user also by display device 670, as
The reference of user's adjusting floating image.The floating image newly adjusted replaces former floating image to be shown in display device 670.Mutual trust
The calculated result of breath computing device 630 can be used as the degree of registration metric of floating image and reference picture.User can also be attached
Show whether floating image is substantially aligned with reference picture by observation with adding.
In the case where determining that floating image and reference picture are substantially aligned, the initial registration position of floating image is obtained,
The position is entered in accuracy registration device 650.It is hung down since the initial registration position based on 3 the accuracy registration position 650
Histogram to sectioning image association relationship, using such as Powell searching algorithm iteratively by floating image and reference picture from
It is dynamic to carry out accuracy registration.The method of the accuracy registration combines Fig. 4 to be described in detail in front.The result of the accuracy registration also by
Display device 670 is shown to user.
Method and apparatus disclosed herein can be applied to clinical auxiliary doctor especially suitable for bone tissue comparative analysis and assessment
Treatment, medical research etc..
The embodiment disclosed herein only describes the method for being registrated to medical threedimensional images by way of example.
It will be understood by those skilled in the art that in the case where not departing from spirit and scope as defined in the appended claims, it can be right
The disclosed embodiments carry out variations and modifications.
Claims (9)
1. a kind of method for being alternatively registrated to medical threedimensional images, comprising:
Medical threedimensional images and reference picture are alternatively subjected to pre-align, in which:
Medicine three is manually adjusted with mutual information of the reference picture on multiple vertical direction in spaces based on medical threedimensional images
Tie up image, and in adjustment process at predetermined intervals and/or adjust number recalculate the mutual information, wherein often
It is secondary that recalculate the mutual information include the mutual information calculated on the multiple vertical direction in space；And
Medical threedimensional images are manually adjusted based on the mutual information recalculated, until obtaining desired initial registration position；With
And
After the pre-align, based on the desired initial registration position by medical threedimensional images and reference picture automatically
Accuracy registration, wherein the accuracy registration is mutual on multiple vertical direction in spaces with reference picture using medical threedimensional images
The value of information is made iteratively.
2. according to the method described in claim 1, wherein manually adjusting medical threedimensional images based on mutual information further includes passing through
The degree of registration of the 3-D image and reference picture on multiple vertical direction in spaces is observed to adjust medical threedimensional images
Registration position.
3. method according to claim 1 or 2, wherein described adjust includes that medical threedimensional images are translated or revolved
Turn, and calculates the mutual information of translation or postrotational medical threedimensional images with reference picture on multiple vertical direction in spaces
Value.
4. according to the method described in claim 1, wherein by medical threedimensional images, automatically accuracy registration includes with reference picture
Determine the association relationship of medical threedimensional images and reference picture on multiple vertical direction in spaces as objective function.
5. according to the method described in claim 4, wherein searching for the optimal solution of the objective function using Powell searching method.
6. according to the method described in claim 5, wherein obtaining search gradient using Parzen window function.
7. method according to claim 1 or 2, plurality of vertical direction in space includes transverse direction, coronal direction
And sagittal direction.
8. method according to claim 1 or 2, wherein the medical threedimensional images are skeletal tissue's 3-D images.
9. a kind of equipment for being alternatively registrated to medical threedimensional images, comprising:
Display device, for showing reference picture and medical threedimensional images to be registered；
Apparatus for adjusting position, for by operator based on medical threedimensional images with reference picture in multiple vertical direction in spaces
Mutual information manually adjust the positions of medical threedimensional images, until obtaining desired initial registration position；
Mutual information computing device, at predetermined intervals and/or adjust number constantly calculate in multiple vertical skies
Between the mutual information on direction and send the mutual information to the apparatus for adjusting position；And
Accuracy registration device for receiving the desired initial registration position from the apparatus for adjusting position, and is based on institute
It states desired initial registration position and utilizes mutual information of the medical threedimensional images with reference picture on multiple vertical direction in spaces
Value is by medical threedimensional images and reference picture automatically accuracy registration.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310746488.8A CN104751438B (en) | 2013-12-31 | 2013-12-31 | For the method and apparatus alternatively to medical threedimensional images registration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310746488.8A CN104751438B (en) | 2013-12-31 | 2013-12-31 | For the method and apparatus alternatively to medical threedimensional images registration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104751438A CN104751438A (en) | 2015-07-01 |
| CN104751438B true CN104751438B (en) | 2019-01-04 |
Family
ID=53591059
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310746488.8A Active CN104751438B (en) | 2013-12-31 | 2013-12-31 | For the method and apparatus alternatively to medical threedimensional images registration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104751438B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105678794B (en) * | 2016-03-01 | 2018-04-17 | 上海电力学院 | A kind of appraisal procedure of the scar treatment effect based on image procossing |
| CN106408509A (en) * | 2016-04-29 | 2017-02-15 | 上海联影医疗科技有限公司 | Registration method and apparatus |
| CN109559296B (en) * | 2018-10-08 | 2020-08-25 | 广州市大智网络科技有限公司 | Medical image registration method and system based on full convolution neural network and mutual information |
| CN109872351A (en) * | 2019-03-18 | 2019-06-11 | 山东浪潮云信息技术有限公司 | A kind of medical figure registration and three-dimensional reconstruction system and method |
| CN113516603B (en) * | 2021-08-03 | 2022-08-19 | 上海交通大学 | Magnetic resonance image multi-modal reconstruction method based on spatial fusion and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100027911A1 (en) * | 2008-08-04 | 2010-02-04 | Siemens Corporate Research, Inc. | Multilevel thresholding for mutual information based registration and image registration using a gpu |
| CN103218811A (en) * | 2013-03-29 | 2013-07-24 | 中国资源卫星应用中心 | Statistical distribution-based satellite multi-spectral image waveband registration method |
| CN103236048A (en) * | 2013-04-18 | 2013-08-07 | 上海交通大学 | Mutual information and interaction-based medical image splicing method |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5449175B2 (en) * | 2007-10-26 | 2014-03-19 | コーニンクレッカ フィリップス エヌ ヴェ | Closed-loop registration control for multi-modality soft tissue imaging |
-
2013
- 2013-12-31 CN CN201310746488.8A patent/CN104751438B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100027911A1 (en) * | 2008-08-04 | 2010-02-04 | Siemens Corporate Research, Inc. | Multilevel thresholding for mutual information based registration and image registration using a gpu |
| CN103218811A (en) * | 2013-03-29 | 2013-07-24 | 中国资源卫星应用中心 | Statistical distribution-based satellite multi-spectral image waveband registration method |
| CN103236048A (en) * | 2013-04-18 | 2013-08-07 | 上海交通大学 | Mutual information and interaction-based medical image splicing method |
Non-Patent Citations (1)
| Title |
|---|
| "医学图像配准和分割技术研究及在图像引导放射治疗系统中的应用";李登旺;《中国博士学位论文全文数据库 信息科技辑》;20111215(第12期);论文第32-53页，图3.7 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104751438A (en) | 2015-07-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104751438B (en) | For the method and apparatus alternatively to medical threedimensional images registration | |
| US8160316B2 (en) | Medical image-processing apparatus and a method for processing medical images | |
| US10282873B2 (en) | Unified coordinate system for multiple CT scans of patient lungs | |
| WO2021238438A1 (en) | Tumor image processing method and apparatus, electronic device, and storage medium | |
| WO2018018981A1 (en) | Method, apparatus and system for re-constructing image of three-dimensional surface | |
| WO2018205232A1 (en) | Method for automatically and accurately positioning reference line according to spliced result | |
| CA3014608C (en) | Systems and methods for characterizing a central axis of a bone from a 3d anatomical image | |
| CN106934821A (en) | A kind of conical beam CT and CT method for registering images based on ICP algorithm and B-spline | |
| CN102222357B (en) | Foot-shaped three-dimensional surface reconstruction method based on image segmentation and grid subdivision | |
| US10548553B2 (en) | Systems and methods for characterizing a central axis of a bone from a 3D anatomical image | |
| CN109767458A (en) | A kind of sequential optimization method for registering of semi-automatic segmentation | |
| CN112561917A (en) | Image segmentation method and system, electronic device and readable storage medium | |
| WO2017028516A1 (en) | Three-dimensional image calibration method, apparatus and system | |
| CN103366348B (en) | A kind of method and treatment facility suppressing skeletal image in X-ray image | |
| WO2024109845A1 (en) | 2d-3d image registration method and system | |
| Pan et al. | Comparison of medical image 3D reconstruction rendering methods for robot-assisted surgery | |
| CN112562070A (en) | Craniosynostosis operation cutting coordinate generation system based on template matching | |
| US20220254104A1 (en) | Detecting and representing anatomical features of an anatomical structure | |
| CN109872353B (en) | White light data and CT data registration method based on improved iterative closest point algorithm | |
| Liu et al. | Optimization of reconstruction of 2D medical images based on computer 3D reconstruction technology | |
| CN101996419B (en) | Three-dimensional reconstruction method of ultrasound image and system thereof | |
| Song et al. | Analysis and Comparison of 3D Reconstruction Volume Rendering Algorithms for Robot-Assisted Surgery | |
| Lu et al. | Multi-scale 2.5 D Integrated Model for False Positive Reduction in Abdominal Lymph Node Detection | |
| Cheng et al. | Endoscopic-based 3D reconstruction methods of human internal cavity | |
| CN113516669A (en) | Trachea extraction method, device and equipment based on CT image and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |