CN1430185A - Ultralarge scale medical image surface reconstruction method based on single-layer surface tracking - Google Patents
Ultralarge scale medical image surface reconstruction method based on single-layer surface tracking Download PDFInfo
- Publication number
- CN1430185A CN1430185A CN 01138698 CN01138698A CN1430185A CN 1430185 A CN1430185 A CN 1430185A CN 01138698 CN01138698 CN 01138698 CN 01138698 A CN01138698 A CN 01138698A CN 1430185 A CN1430185 A CN 1430185A
- Authority
- CN
- China
- Prior art keywords
- described method
- data
- medical image
- tracking
- surface tracking
- 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
Images
Abstract
A method based on single-layer surface tracking for reconstructing the surface of very-large-scale medical image includes such steps as extracting the interesting part from 3D method image, extracting the interesting organ surface by single-layer surface tracking techinque, processing the extracted 3D surface model for fast drawing, and real-time display and interaction to 3D surface model of organ. Its advantages are use of PC and Windows operating system, low cost, easy operation, less space requirement to memory, and high speed.
Description
Technical field
The present invention relates to pattern-recognition, particularly based on the ultra-large medical image method of surface reconstruction of individual layer surface tracking.
Prior art
Since the X ray invention, modern medicine image documentation equipments such as CT, MRI, CR, B ultrasonic, fujinon electronic video endoscope successively occur, and make traditional medical diagnosis mode that revolutionary variation take place.The appearance of medical image has brought new dawn to the research of medical diagnosis, modern computer science and technology development particularly, and making becomes the interdisciplinary science field of an emerging development based on the diagnosis of medical image.The hospital of China between more than ten years, has introduced large quantities of advanced persons' medical image equipment in the past, has played important positive role to improving diagnostic level.
But, all be two-dimentional by the resulting image of medical imaging device, need trained radiologist to make judgement.Along with the development of computer visualization technology, by assisting of computing machine, can generate 3-D view true to nature with a series of two-dimensional medical images through handling, make the doctor see better, see more accurately, see more conveniently.
The three-dimensional visualization of medical image is that the core of this subject is formed, and it utilizes a series of two-dimensional slice image to carry out reconstructing three-dimensional model and demonstration, is the prerequisite of carrying out quantitative test.Two kinds of rendering techniques during realizing, three-dimensional visualization are arranged: surface rendering and direct volume drawing.The characteristics of iso-surface patch maximum are to need to advance 2-dimension data field earlier that objects is cut apart and three-dimensional reconstruction, and the representation of a surface of formation object edge equivalence adopts the illumination model drawing image again.And volume drawing is to regard the voxel in the three-dimensional data as vitrina one by one, and classifies and give its certain color and opacity, passes whole data fields by light, carries out color and synthesizes, and obtains final drawing result.
Because surperficial display speed is fast, and can provide the tissue more true to nature and the interactive mode demonstration of organ for the doctor, thereby improve the accuracy of medical diagnosis, so resurfacing and surperficial display technique are used more and more widely in Medical Image Processing and in analyzing.Wherein famous is exactly MC (Marching Cubes) algorithm that was put forward by W.Lorensen in 87 years, because the MC algorithm is realized simple, and can obtain the output of high-res, can also make full use of the graphic presentation hardware of current main-stream, therefore make it become most popular surface extraction algorithm.But the MC algorithm will generate a large amount of tri patchs and remove to approach the target curved surface when resurfacing, and this makes must handle relatively difficulty of extensive model, and in addition, the speed of MC algorithm is slow.Although the researchist has developed a large amount of methods and solved these problems, these methods or calculating that need be extra perhaps need a large amount of auxiliary storage spaces, all do not fit into the real-time processing of ultra-large data set.
Some commercial softwares have been arranged at present, as the AVS of MEDx, the Advanced Visual Systems of ANALYZE, the SensorSystems of Mayo Clinic and Stealth Station of pivot Fa Modanli etc., they provide all that a large amount of instruments is used for carrying out 2D, 3D rendering demonstration, Flame Image Process, cuts apart, registration, quantitative test etc., and surperficial Presentation Function all is provided.But because the calculated amount of three-dimensional reconstruction is very big, these business softwares often require to operate on the high-grade workstation, and cost an arm and a leg, so range of application has been received very big restriction.
Because the present image resolution ratio that medical imaging device produced more and more higher (as 1024 * 1024), data are more and more, also brought bigger challenge for three-dimensional reconstruction.Particularly since the virtual human body project of the U.S. occurred, the data of magnanimity made processing in real time and demonstration become more difficult, thereby made that the new method that can efficiently handle ultra-large medical images data sets of exploitation is particularly important.
Summary of the invention
Thereby the purpose of this invention is to provide a kind of method of surface reconstruction that ultra-large medical images data sets obtains the organ surface shape of high-res of efficiently handling apace, this method is target platform with the common PC, handles fast and efficiently to have high-resolution ultra-large medical image.
For achieving the above object, the ultra-large medical image method of surface reconstruction based on the individual layer surface tracking comprises:
(1) segmentation procedure splits interested part from the two dimension slicing of medical image;
(2) surface extraction step uses the individual layer surface tracking that interested organ surface is extracted;
(3) V-belt generates step, and the 3 d surface model that extracts is handled, and makes it be suitable for quick drafting;
(4) interactive step display carries out the sense of reality to the 3 d surface model of organ and shows and real-time, interactive.
Target platform of the present invention is at the very high common PC of China's popularity rate, and operating system is the Windows series operating system of friendly interface, and is not only with low cost, and operation easily.By surface tracking and the geometric data compress technique of utilizing individual layer, significantly reduced the memory consumption of algorithm, also guaranteed very fast reconstruction speed simultaneously.In medical image amount increasing today, this invention has important use at medical domain and is worth, and has high confidence level, applicability and admissibility.
Description of drawings
Fig. 1 is based on the pie graph of the ultra-large medical image method of surface reconstruction of individual layer surface tracking;
Fig. 2 is to use the connectivity synoptic diagram that can correctly handle based on the individual layer surface tracking algorithm;
In Fig. 2, represent with thick line and what represent with fine rule is two parts of a continuous curve surface, but their major parts in one deck section are disconnected.The method that will describe below using, they finally have been connected in the surface.
Fig. 3 is to use the connectivity synoptic diagram that can correctly handle based on the individual layer surface tracking algorithm;
In Fig. 3, represent with thick line and what represent with fine rule is two parts of a continuous curve surface, but their major parts in one deck section are disconnected.By the thresholding method that use describes below, they finally can be connected in the surface.
Fig. 4 is that list item is the synoptic diagram how to organize in the look-up table;
In Fig. 4 (a), enumerated the example of a simple voxel configuration, (b) figure is the synoptic diagram of six direction, (c) figure has provided the value of the list item of its correspondence in look-up table.
Fig. 5 is the synoptic diagram of used data buffer;
Wherein, Fig. 5 (a) has illustrated the position (Top Edges, Bottom Edges and Z Edges) of three buffer zones, and (b) sectional view of data set has been illustrated on the figure left side, and how the content in the right signal buffer zone is to bringing Forward.
Fig. 6 is the signal of a triangular surface patch grid and adjacent map thereof;
Fig. 7 is that the bone of virtual human body is rebuild;
Fig. 8 is that skin is rebuild design sketch.
The working of an invention mode
At present PC is constantly popularized in China, and price also constantly descending, and the hardware capability of PC is with exponential development by leaps and bounds.Novel high-speed CPU constantly occurs, and has reached GHz now, and processing speed is more and more faster, and constantly has new instruction set the execution of multimedia application to occur quickening.Support the video card frequent updating of three-dimensional hardware-accelerated technology to regenerate, occurred now than CPU processing speed graphic process unit (GPU) faster, as GeForce2 of NVidia etc., can use the function that hardware is finished originally needed software to finish such as three-dimensional picture conversion, illumination, cutting etc.All these conditions make that all developing the practical method that can efficiently handle ultra-large data set on current main flow PC becomes possibility.
Describe the resurfacing and the display packing of ultra-large medical image of the present invention below in detail.This implementation is made up of four key steps, and structural drawing can be referring to Fig. 1.These four steps are respectively: segmentation procedure, surface extraction step, Triangle Strips generate step and interactive step display, are introduced one by one below.
Cut apart
The purpose in this step is to do pre-service for the resurfacing algorithm, and target object is split from background, is also referred to as the process of binaryzation.It is vital cutting apart for high-quality three-dimensional reconstruction, because it is determining that whether the object that finally shows is our interested organ.
Dividing method has a lot of different kinds, and each kind all is suitable for different source images.More effective such as Threshold Segmentation to CT, but for the MRI image, because inside of human body complex structure, the wriggling of biological tissue and the characteristics of MRI imaging, cause that target object inevitably is subjected to other object or even interference of noise in the medical image, make object local edge feature fuzzy, just be difficult to obtain effect preferably with Threshold Segmentation.So the best way combines dividing method and three-dimensional rebuilding method exactly, dividing method as much as possible is provided, select different dividing methods for use at different source images, the segmentation result that obtains pinpoint accuracy is application surface method for reconstructing more later on.
Here we introduce two kinds of practical dividing methods: threshold value method and region growing method.The key of threshold method is the selection of threshold value, can be selected to distinguish the gray threshold of background and non-background by the user, and also available automatic threshold method is determined threshold value.Common automatic threshold method has the P-parametric method, state method, differential histogram method, techniques of discriminant analysis and variable thresholding method.At the many characteristics of medical image noise, can adopt techniques of discriminant analysis.Promptly in the histogram of gradation of image value, try to achieve threshold value t the set of gray-scale value is divided into two groups, make two groups to obtain optimal separation.The standard of optimal separation is that the ratio of two groups the variance of mean value and each prescription difference is for maximum.When this method had two crests in histogram, the state method of can be used as worked; Even also can obtain threshold value when not having crest.If given image has L level gray-scale value, threshold value is k, and k is divided into two group 1,2 with the pixel of image.The pixel count of group 1 is made as ω
1(k), average gray value is M
1(k), variance is σ
1(k)); The pixel count of group 2 is made as ω
2(k), average gray value is M
2(k), variance is σ
2(k).If the average gray value of all pixels is decided to be M τ.Then:
Variance in the group
Variance between group
Optimality criterion
Value is for maximum, promptly
Get maximal value.
For the region growing method, need the user to select a point on the contoured skin as seed points.The basic thought of region growing is that the pixel collection that will have similar quality gets up to constitute the zone, and this method need be chosen a seed points earlier, will plant subpixel similar pixel on every side then successively and merge in kind of the zone at subpixel place.The research emphasis one of region growing algorithm is the design of characteristic measure and region growing rule, the 2nd, and the high efficiency of algorithm and accuracy.We use the symmetrical region growth algorithm, and can remedy two big weakness of region growing algorithm effectively: to the selection sensitivity of initial seed point, and EMS memory occupation is too much, and to 3D connecting object mark and the empty efficiency of algorithm height of deletion.
Surface extraction
To in large-scale dataset, reconstruct interested three-dimensional surface apace, have several factors to consider in earnest:
1) memory consumption.Because our target platform is a common PC, for resembling the so large-scale data set of virtual human body, the reasonable use of internal memory is very important.
2) traversal speed.Because voxel number to be processed is too many, speed how to quicken to travel through voxel also is very crucial.
3) render speed.The a large amount of tri patch that produces will bring very big challenge to render speed.
But very unfortunate, in these factors, the consumption of internal memory and the speed of traversal are a pair of paradox.If use some space segmentation technology such as Octree to wait the traversal speed of accelerating voxel, then except data set, in internal memory, also to store supplementary structures such as Octree, this obviously is unpractiaca for large-scale dataset; If the saving internal memory then must sequentially travel through voxel, this will reduce the speed of rebuilding.
Here the algorithm synthesis that is proposed has been considered above factor, and it can reconstruct interested surface, and show in real time under the situation that consumes little memory apace.
Because the surface tracking at six direction makes data buffering become difficult, and need more auxiliary storage space, so proposed the surface tracking technology of individual layer here, promptly only the four direction on section plane, place carries out surface tracking, with the perpendicular direction of section on handle according to order from top to bottom, our efficient that experimental results show that this algorithm is quite high.
In the Marching of standard Cubes algorithm, the order of voxel traversal is from left to right, from top to bottom, does not consider the neighbor information of voxel, so the most of the time of algorithm flower is in the detection to empty voxel.If use traditional surface tracking algorithm, then need whole data set is read in internal memory, and need carry out random access whole data set, this has limited its application on large-scale dataset.
Our thought is both to have utilized the advantage of surface tracking, lacks consumes memory again as far as possible, and realizes data set according to the section sequential access, thereby obtains a best equilibrium point on speed and memory consumption.Very naturally,, we are limited in surface tracking in the individual layer, thereby have reached this purpose.
In the algorithm that here puts forward, slice of data is read in internal memory according to order from top to bottom by burst, and one deck has been formed in per two sections.Algorithm only carries out surface tracking at four direction in one deck, read into to descend a slice slice of data again after handling one deck.
The individual layer surface tracking is compared with the surface tracking of six direction, may cause a problem: the surface that extracts not exclusively.Because lack the degree of freedom of both direction, be that interconnective curved surface may not connect in a section in three dimensions, as shown in Figure 2.In order to address this problem, we adopt sequential scanning in the processing of ground floor, to upwards there be the cube of connection to join in the seed points set simultaneously up, when one deck is cut into slices under handling, just from this seed points set, in the enterprising line trace of four direction, same, be recorded in the cube that upwards there is connection the top.Propagate by such seed points, not only improved arithmetic speed, also can partly address the above problem.In Fig. 2, the part of representing with thick line and represent with fine rule is not communicated with in the major part section, but by this method, finally has been connected in the surface.But even now has still lost a downward degree of freedom, and when not comprising all seed points in the ground floor, algorithm still may only search out part surface, as shown in Figure 3, has only the part of representing with thick line to be extracted out.Although at this moment can search for downwards, a part of slice of data will be repeated to read again, and the complexity of whole algorithm also greatly increases simultaneously.Here we have used a simple method, a tri patch number thresholding is set,, then in this layer, carries out sequential scanning again if the tri patch number that extracts in certain one deck is less than this thresholding, obtain seed points complete in this layer, and then upwards propagate.Facts have proved that suitable if thresholding is selected, it is more effective doing like this.
In order in single layer data, to carry out surface tracking fast, can construct a look-up table NeighbourTable, it all uses a byte to write down link information with neighbours' voxel to 256 kinds of possible voxel configuration modes.The Senior Three position representative of byte effective number of voxel in this voxel neighbours voxel, scope is 0-6, low five from a high position to the low level, represent respectively, whether the neighbours' voxel on five directions in front, rear, left and right effective, wherein 1 representative effectively, 0 represent invalid.For example, for voxel shown in Figure 4, the list item in its pairing NeighbourTable should be 01111001.
The number of considering the output tri patch in addition is very huge, if adopt the traditional summit table and the grid expression way of face table, will expend a large amount of internal memories.Here for the tissue of vertex data, we have taked a kind of compact data structure.If the resolution of data set is Ix * Iy * Iz, and suppose that maximum one dimension is Ix, at first distribute Iy * Iz pointer PointsList
*, wherein the PointsList structure organization is as follows:
X coordinate: 16bit
Normal vector: 16bit
Normal vector is quantized on a unit cube, and each face is divided into 100 * 100 points, and 6 * 100 * 100 different normal vectors are quantized into 16bit altogether, and such precision can't cause decrease in image quality visually.
After adopting this data structure, the data of preserving M point in the grid only need Iy * Iz * 4+M * 4 byte, if and use traditional coordinate method vector representation just to need the byte of M * (3+3) * 4, from our observation, under most of situation new method only take classic method less than 1/4 storage space.
Algorithm at first takes out a point from the seed points set that the upper strata obtains when one deck is followed the tracks of, if this point does not have accessed mistake, then it is joined a rear of queue.Algorithm all goes the head of formation to take out a voxel at every turn, obtains the intersection point (point coordinate in directly adopting) on the interior contour surface of this voxel and each limit herein, and obtains the connected mode of triangular plate.Obtain the link information of this voxel on four direction by searching the NeighbourTable table then, the neighbours' voxel for being connected if it does not have accessed mistake, then joins rear of queue with it.Will judge also simultaneously whether this voxel has to connection up, if having, record advances in the seed points set.Process above circulation is carried out, when formation was sky, the contour surface that is connected in the individual layer just had been extracted out.
In the process of surface tracking, can utilize data Caching Mechanism to avoid the calculating of some repetitions.Say usually that for an individuality it and each neighbours' voxel all will be shared a face, the equivalent point on these faces in fact only needs to calculate once, is kept in the data buffer zone, and just can directly directly take out result of calculation from buffering next time.The difficult point that realizes a reasonable data buffering algorithm just is to reduce the shared internal memory in data buffer how as far as possible, and will when data are otiose it be discharged, and uses for other data.
Because this algorithm is limited in the individual layer the inside with surface tracking, so can realize a kind of active data buffering at an easy rate.Used three buffer zones to write down top layer limit (TopEdges), lower floor limit (Bottom Edges) and middle limit (Z Edges) respectively here, as shown in Figure 5.The height and width of supposing slice of data are respectively ImageHeight and ImageWidth, then the length of top layer limit and lower floor's limit buffer zone all is 2 * ImageHeight * ImageWidth, and medial side only needs the length of ImageHeight * ImageWidth, so not only having guaranteed all has a hash table institute corresponding to every limit, simultaneously also without any the waste on the internal memory.
For the expression of contour surface, the traditional summit table and the expression way of face table have been adopted here.Summit Table V ertexTable has write down the coordinate and the normal vector on all summits, and its expression-form is as follows:
{x，y，z；nx，ny，nz}
X wherein, y, z are three coordinates on summit, nx, ny, nz are the normal vector of summit on three directions.And face table FacetTablei and traditional expression way are different, and it has only write down the connected mode of the contour surface in one deck, and its expression-form is as follows:
{index1，index2，index3}
Index1 wherein, index2, index3 represent three location indexs of summit in VertexTable of a tri patch respectively.After the contour surface of one deck was extracted out, FacetTablei just was added in the chained list and goes.
All list items of three buffer zones (Top Edges, Bottom Edges and Z Edges) the inside all are changed to Empty under the most initial state, when handling certain bar limit of a voxel, if equivalent point is arranged on this edge, so at first go to search in the corresponding buffer region whether corresponding locational list item is Empty, if be Empty, will calculate coordinate, the normal vector of this point so, and be inserted among the Table V ertexTable of summit, at this moment also to upgrade corresponding list item simultaneously, write down the index of this summit in VertexTable; If list item is not Empty, represent that this summit calculated in front, can directly take out the vertex index of the inside record, join among the FacetTablei and get final product.
After having calculated the contour surface of one deck, the content of the buffer zone the inside one deck of also should boosting can be seen buffer zone as a window, the whole lattice that up moved of this window.At first, give ground floor the content assignment of the buffer zone the inside second layer; Be Empty clearly all with all contents in table in the second layer then, so just reused the space of limited data buffering, as shown in Figure 5.
V-belt (Triangle Strips) generates
By two top steps, the triangular surface patch grid that has obtained contour surface is expressed.Because a tri patch could be expressed in three summits,, 3 * n summit just must be arranged so want to express grid with n tri patch; And if use V-belt (Triangle Strips) because shared the summit between the tri patch, so can only use n+2 summit just can express n tri patch.Can save transfer bandwidth between the internal memory of internal memory and display card so in large quantities, greatly improve display speed.
In order to draw the grid that generates more quickly, adopted a V-belt generating algorithm fast here, the contour surface that extracts is generated compacter V-belt express.Because the top data Caching Mechanism that adopted, thereby guarantee not have the summit of repetition in the Table V ertexTable of summit, and what write down in face table FacetTablei is the index value on summit, algorithm can only be operated in the face table like this.Be the description of algorithm below.
At first, algorithm is created out an adjacent map by the face table.In fact so-called adjacent map is exactly a big Hash table, and it has write down the neighbor information of each triangular plate on every limit, and Fig. 6 has shown a triangle gridding and its pairing adjacent map.Create adjacent map, in face table FacetTablei, search the tri patch that all share a limit, and they are inserted in the adjacent map.
Secondly, obtain after the adjacent map, from figure, find have the minimum number of degrees triangular plate of (neighbours' number).If there are several triangular plates all to have the same number of degrees, search their neighbours' the number of degrees so again, choose that triangular plate of neighbours' number of degrees minimum; If their neighbours' the number of degrees are still the same, that can at will choose one.
Then, from this triangular plate of choosing, circulation generates the V-belt.Selected triangular plate is joined in the V-belt, and search the triangular plate with minimum number of degrees in its neighbours, it is added in the V-belt, circulation so always is till all accessed mistake of all neighbours.If run into a triangular plate, it should create a new V-belt so without any neighbours this moment; If when having run into the triangular plate that several identical number of degrees are arranged, disposal route and top introduce identical.For fear of infinite loop, after a triangular plate is joined the V-belt, need in adjacent map, it and all pointers that point to it all be removed.
Whether at last, check whether the V-belt that generates has organized all summits effectively, be effective V-belt.At this moment to detect all triangular plates in the V-belt, see whether its latter two summit is preceding two summits of next triangular plate.
The interactive demonstration
After obtaining extracting the three-dimensional model of organ, how their interactivelies being shown also is a considerable problem.In order to excavate the potential of ordinary PC as much as possible, realize that the high-speed sense of reality shows that we have considered following factor:
1) uses ripe three-dimensional picture API OpenGL, utilize hardware-accelerated
OpenGL is the three-dimensional picture API of present comparative maturity, it can use on the current PC the hardware-accelerated technology that provided of the video card (as the GeForce series of nVidia company) of a large amount of main flows that adopt, become the industrial standard of current 3-D display API aspect.
Because we have been expressed as three-dimensional model the form of V-belt (TriangleStrips) in a last step, and present main flow video card has all been done optimization at the demonstration of V-belt specially, so we have used OpenGL expansion NV_vertex_array_range and the NV_fence of nVidia, from our experiment, on the machine of a PIII800, probably can draw about 7,000,000 tri patchs one second, can satisfy the real-time display requirement of most of large-scale data.
2) realize continuous LoD control with software, software and hardware combining realizes real-time rendering
For the so ultra-large data set of virtual human body, the three-dimensional model that extracts may include up to ten million even more than one hundred million tri patchs, if pure is drawn with the ability of OpenGL, still can not get satisfied result.So must realize LoD control and blanking with software, this will alleviate the work load of the GPU of video card greatly, and can utilize the arithmetic capability of CPU and GPU cmpletely, make their concurrent workings simultaneously.
For the triangular surface patch grid that extracts, it is organized in the border ball of an into layering.In drawing process, use layering border ball to carry out visuality judgement, LoD control.By traveling through layering border ball from top to bottom, if a border ball is visible, and project to the thresholding that area on the screen is lower than a setting, this border ball is drawn so.When the user uses the mutual control of mouse,, the area thresholding can be established higher a little in order to safeguard one than higher frame rate; System's free time, the area thresholding can be made as a pixel, obtain the image of E.B.B., therefore can realize LoD control very naturally.
By using above-described software and hardware blend rendering method, handle ultra-large data set, can on common PC, just can obtain good effect.
Embodiment
We apply to us with the method and design voluntarily in the 3 d medical images processing and analytic system of realization.The 3 d medical images Treatment Analysis 3dmed of system based on microcomputer that we develop is under microcomputer Windows NT and Windows98 environment, adopt Object Oriented method and soft project standard, realize with C Plus Plus, handle and analytic system towards the 3-D view of medical domain.Native system has abundant graph and image processing and analytic function, not only has perfect two dimensional image Treatment Analysis function, and has powerful 3D processing and functions such as analysis, Network Transmission and storage.The function that system provides comprises a series of functions such as data input, image data management, two-dimensional process, three-dimensional data processing, section reorganization, 3-D display, surgical simulation, virtual endoscope, PACS and remote diagnosis.
The following describes utilization and handle the specific implementation process of ultra-large medical images data sets based on the method for individual layer surface tracking.The virtual human body data set of experimental data for providing on the u.s. national library of medicine website, used herein is the CT data set, the data set scale is 512 * 512 * 522.Because the difference of some parameters is (such as fov when data are obtained, pixel size, slice spacings etc.), whole data set has been divided into nine parts, because the interval between last part (data of foot) and the last partial data is too big, we have only used the first eight part.The concrete operations step is as follows:
1) at first by the data-interface reading of data.
2) click " senior structure " button, enter and cut apart the interface.
3) it is available that system provides multiple dividing method, has seed growth, corrosion to expand, blur connection, thresholding, Interactive Segmentation etc., can choose a kind of dividing method this moment data are cut apart.Because what handle is the CT data, the thresholding method is more effective, specify low thresholding and high thresholding with mouse after, system will split the material within this thresholding.
4) cut apart after, point " 3D demonstration " button, the algorithm that system will call described in the application carries out three-dimensional reconstruction, and the photo realism graphic after will rebuilding shows, and allows the user to use mouse to carry out interactive mode observation.We have rebuild skin and two models of bone respectively to the virtual human body data, and design sketch as shown in Figure 7 and Figure 8.
Above-mentioned experimental result is consistent with the theoretical analysis conclusion that the inventor handles ultra-large medical images data sets to utilization based on the surface tracking of individual layer, has high confidence level, applicability and admissibility.
Claims (8)
1. ultra-large medical image method of surface reconstruction based on the individual layer surface tracking comprises:
(1) segmentation procedure splits interested part from the two dimension slicing of medical image;
(2) surface extraction step uses the individual layer surface tracking that interested organ surface is extracted;
(3) V-belt generates step, and the 3 d surface model that extracts is handled, and makes it be suitable for quick drafting;
(4) interactive step display carries out the sense of reality to the 3 d surface model of organ and shows and real-time, interactive.
2. by the described method of claim 1, it is characterized in that described individual layer surface tracking comprises step:
Four direction on section plane, place carries out surface tracking;
With the perpendicular direction of section on from top to bottom order handle.
3. by the described method of claim 2, it is characterized in that also comprising step:
In the processing of ground floor, adopt sequential scanning;
To upwards there be the cube of connection to join in the seed points set simultaneously up;
When handling next section, from this seed points set, in the enterprising line trace of four direction;
The cube that connection is upwards arranged above being recorded in.
4. by the described method of claim 3, it is characterized in that comprising step:
Tri patch number threshold value is set,, then in this layer, carries out sequential scanning again, obtain seed points complete in this layer, and then upwards propagate if the tri patch number that extracts in certain one deck is less than threshold value.
5. by the described method of claim 2, it is characterized in that described surface tracking comprises buffer zone is set, be used for writing down respectively top layer limit, lower floor limit and middle limit.
6. by the described method of claim 2, it is characterized in that described buffer zone has three.
7. by the described method of claim 2, it is characterized in that the grid compression expression that is adopted:
If the resolution of data set is Ix * Iy * Iz, and suppose that maximum one dimension is Ix, then at first distribute Iy * Iz pointer PointsList.
8. by the described method of claim 7, it is characterized in that described PointsList data structure is:
X coordinate: 16bit;
Normal vector: 16bit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011386983A CN100541541C (en) | 2001-12-29 | 2001-12-29 | Ultra-large medical image method of surface reconstruction based on the individual layer surface tracking |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB011386983A CN100541541C (en) | 2001-12-29 | 2001-12-29 | Ultra-large medical image method of surface reconstruction based on the individual layer surface tracking |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1430185A true CN1430185A (en) | 2003-07-16 |
| CN100541541C CN100541541C (en) | 2009-09-16 |
Family
ID=4674668
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB011386983A Expired - Fee Related CN100541541C (en) | 2001-12-29 | 2001-12-29 | Ultra-large medical image method of surface reconstruction based on the individual layer surface tracking |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100541541C (en) |
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1296874C (en) * | 2003-12-22 | 2007-01-24 | 中国科学院自动化研究所 | Point reconstruction based very large-scale medical image 3D visualization method |
| CN1301494C (en) * | 2004-06-07 | 2007-02-21 | 东软飞利浦医疗设备系统有限责任公司 | Three-dimensional dividing method for medical images |
| CN1315102C (en) * | 2003-12-12 | 2007-05-09 | 西北工业大学 | Method and system for transparent roaming body of object |
| CN101189639B (en) * | 2005-04-13 | 2010-09-29 | 皇家飞利浦电子股份有限公司 | A method and a system for segmenting a surface in a multi-dimensional dataset |
| CN101976465A (en) * | 2010-10-27 | 2011-02-16 | 浙江大学 | Acceleration improvement algorithm based on cube edge sharing equivalent point |
| CN101116104B (en) * | 2005-02-10 | 2012-07-04 | 皇家飞利浦电子股份有限公司 | A method, and a system for segmenting a surface in a multidimensional dataset |
| CN101710420B (en) * | 2009-12-18 | 2013-02-27 | 华南师范大学 | Anti-segmentation method for medical image |
| CN103413349A (en) * | 2013-07-19 | 2013-11-27 | 东北大学 | DICOM image three-dimensional visualization method based on IOS |
| CN104933751A (en) * | 2015-07-20 | 2015-09-23 | 上海交通大学医学院附属瑞金医院 | Angiocarpy coronary artery enhanced volume rendering method and system based on local histograms |
| CN106127750A (en) * | 2016-06-20 | 2016-11-16 | 中国科学院深圳先进技术研究院 | A kind of CT image body surface extracting method and system |
| CN106296756A (en) * | 2015-06-04 | 2017-01-04 | 高德软件有限公司 | A kind of polygonal compression method and device |
| CN106844770A (en) * | 2017-02-28 | 2017-06-13 | 成都金盘电子科大多媒体技术有限公司 | A kind of three-dimensional modeling imaging optimization method based on medical image |
| CN107146222A (en) * | 2017-04-21 | 2017-09-08 | 华中师范大学 | Medical Image Compression algorithm based on human anatomic structure similitude |
| CN107278316A (en) * | 2016-07-14 | 2017-10-20 | 中山大学附属第医院 | Beamforming structure three-dimensional rebuilds visualization integrated approach inside human peripheral nerve |
| WO2018001099A1 (en) * | 2016-06-30 | 2018-01-04 | 上海联影医疗科技有限公司 | Method and system for extracting blood vessel |
| CN108961281A (en) * | 2018-03-28 | 2018-12-07 | 研靖信息科技（上海）有限公司 | A kind of image partition method and equipment based on 3D voxel data image |
| CN108986111A (en) * | 2018-07-02 | 2018-12-11 | 西安增材制造国家研究院有限公司 | A kind of three-dimensional dentognathic model dividing method for area of computer aided stealth correction |
-
2001
- 2001-12-29 CN CNB011386983A patent/CN100541541C/en not_active Expired - Fee Related
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1315102C (en) * | 2003-12-12 | 2007-05-09 | 西北工业大学 | Method and system for transparent roaming body of object |
| CN1296874C (en) * | 2003-12-22 | 2007-01-24 | 中国科学院自动化研究所 | Point reconstruction based very large-scale medical image 3D visualization method |
| CN1301494C (en) * | 2004-06-07 | 2007-02-21 | 东软飞利浦医疗设备系统有限责任公司 | Three-dimensional dividing method for medical images |
| CN101116104B (en) * | 2005-02-10 | 2012-07-04 | 皇家飞利浦电子股份有限公司 | A method, and a system for segmenting a surface in a multidimensional dataset |
| CN101189639B (en) * | 2005-04-13 | 2010-09-29 | 皇家飞利浦电子股份有限公司 | A method and a system for segmenting a surface in a multi-dimensional dataset |
| CN101710420B (en) * | 2009-12-18 | 2013-02-27 | 华南师范大学 | Anti-segmentation method for medical image |
| CN101976465A (en) * | 2010-10-27 | 2011-02-16 | 浙江大学 | Acceleration improvement algorithm based on cube edge sharing equivalent point |
| CN103413349A (en) * | 2013-07-19 | 2013-11-27 | 东北大学 | DICOM image three-dimensional visualization method based on IOS |
| CN106296756A (en) * | 2015-06-04 | 2017-01-04 | 高德软件有限公司 | A kind of polygonal compression method and device |
| CN104933751B (en) * | 2015-07-20 | 2017-10-20 | 上海交通大学医学院附属瑞金医院 | The enhanced object plotting method of cardiovascular coronary artery and system based on local histogram |
| CN104933751A (en) * | 2015-07-20 | 2015-09-23 | 上海交通大学医学院附属瑞金医院 | Angiocarpy coronary artery enhanced volume rendering method and system based on local histograms |
| CN106127750B (en) * | 2016-06-20 | 2019-07-30 | 中国科学院深圳先进技术研究院 | A kind of CT images body surface extracting method and system |
| CN106127750A (en) * | 2016-06-20 | 2016-11-16 | 中国科学院深圳先进技术研究院 | A kind of CT image body surface extracting method and system |
| US10357218B2 (en) | 2016-06-30 | 2019-07-23 | Shanghai United Imaging Healthcare Co., Ltd. | Methods and systems for extracting blood vessel |
| WO2018001099A1 (en) * | 2016-06-30 | 2018-01-04 | 上海联影医疗科技有限公司 | Method and system for extracting blood vessel |
| US11344273B2 (en) | 2016-06-30 | 2022-05-31 | Shanghai United Imaging Healthcare Co., Ltd. | Methods and systems for extracting blood vessel |
| CN107278316A (en) * | 2016-07-14 | 2017-10-20 | 中山大学附属第医院 | Beamforming structure three-dimensional rebuilds visualization integrated approach inside human peripheral nerve |
| CN106844770A (en) * | 2017-02-28 | 2017-06-13 | 成都金盘电子科大多媒体技术有限公司 | A kind of three-dimensional modeling imaging optimization method based on medical image |
| CN107146222A (en) * | 2017-04-21 | 2017-09-08 | 华中师范大学 | Medical Image Compression algorithm based on human anatomic structure similitude |
| CN107146222B (en) * | 2017-04-21 | 2020-03-10 | 华中师范大学 | Medical image compression method based on human anatomy structure similarity |
| CN108961281A (en) * | 2018-03-28 | 2018-12-07 | 研靖信息科技（上海）有限公司 | A kind of image partition method and equipment based on 3D voxel data image |
| CN108986111A (en) * | 2018-07-02 | 2018-12-11 | 西安增材制造国家研究院有限公司 | A kind of three-dimensional dentognathic model dividing method for area of computer aided stealth correction |
| CN108986111B (en) * | 2018-07-02 | 2021-08-17 | 西安增材制造国家研究院有限公司 | Three-dimensional dental model segmentation method for computer-aided invisible orthodontics |
Also Published As
| Publication number | Publication date |
|---|---|
| CN100541541C (en) | 2009-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100541541C (en) | Ultra-large medical image method of surface reconstruction based on the individual layer surface tracking | |
| US4737921A (en) | Three dimensional medical image display system | |
| CN1271572C (en) | Three dimension re-set-up method for two dimension image sequence | |
| Jeong et al. | Ssecrett and neurotrace: interactive visualization and analysis tools for large-scale neuroscience data sets | |
| CN103106685B (en) | A kind of abdominal organs three-dimensional visualization method based on GPU | |
| CN101625767B (en) | Method for rebuilding point-cloud type three-dimensional surface of nonparallel outline medical image | |
| Goldwasser et al. | Real-time display and manipulation of 3-D medical objects: The voxel processor architecture | |
| CN101324962A (en) | Parallel processing method drawn by pre-projection light ray projection body | |
| US20090079736A1 (en) | Information processing apparatus and program | |
| CN106570930A (en) | Digital biological organism modeling method based on three-dimensional super voxel description | |
| CN108537750A (en) | A kind of quick spine CT 3-dimensional reconstruction method | |
| CN1296874C (en) | Point reconstruction based very large-scale medical image 3D visualization method | |
| CN1403057A (en) | 3D Euclidean distance transformation process for soft tissue display in CT image | |
| CN109360219A (en) | A kind of augmented reality auxiliary operation method and system | |
| Noon | A volume rendering engine for desktops, laptops, mobile devices and immersive virtual reality systems using GPU-based volume raycasting | |
| CN103745495A (en) | Medical volume data based volume rendering method | |
| CN1779718A (en) | Visula partitioned drawing device and method for virtual endoscope | |
| CN1315102C (en) | Method and system for transparent roaming body of object | |
| CN108830922A (en) | A kind of profile tree constructing method based on multithreading | |
| Yagel | Classification and survey of algorithms for volume viewing | |
| US11955246B2 (en) | Unified anisotropic volume and surface mesh storage | |
| Shin et al. | A real-time display methods for large-scale human body data | |
| Yazdy et al. | Carvupp: computer assisted radiological visualisation using parallel processing | |
| Yagel | Volume viewing algorithms: Survey | |
| CN2787139Y (en) | Visible divisional drawing device of virtual endoscope |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090916 Termination date: 20111229 |