CN100481128C - Method and system for using cutting planes for colon polyp detection - Google Patents
Method and system for using cutting planes for colon polyp detection Download PDFInfo
- Publication number
- CN100481128C CN100481128C CNB2004800272532A CN200480027253A CN100481128C CN 100481128 C CN100481128 C CN 100481128C CN B2004800272532 A CNB2004800272532 A CN B2004800272532A CN 200480027253 A CN200480027253 A CN 200480027253A CN 100481128 C CN100481128 C CN 100481128C
- Authority
- CN
- China
- Prior art keywords
- cutting planes
- image
- voxel
- colon
- intersection
- 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.)
- Expired - Fee Related
Links
Images
Abstract
A method of identifying a polyp in a digital image of a colon, wherein said image comprises a plurality of intensities corresponding to a domain of voxels in a 3 dimensional space, is provided. The method includes providing the image with a set of 3 mutually orthogonal axes, providing a plurality of cutting planes each at a different orientation with respect to the image axes, centering (101), for each voxel in the image, each of the cutting planes about a central voxel, determining, for each of the plurality of cutting planes about each voxel in the image, an intersection of the cutting plane with the colon, and examining (102) a trace of the cutting plane within said intersection, and marking (103), where the trace of each cutting plane is small and round, those voxels in the intersection for further analysis.
Description
The cross reference of relevant U. S. application
The application requires to be called by the name that Pascal Cathier submitted on September 22nd, 2003 right of priority of the U.S. Provisional Application No.60/504714 of " Cutting planes for colon polyp detection ", and the content of this application is merged in by reference at this.
Technical field
The present invention relates to structure, especially polyp of colon in the discriminating digit medical image.
Background technology
Allow in early days and the stage that more can treat detection potential problems from the obtainable diagnosis high-level information of the data of obtaining by current imaging system.If given a large amount of, must develop then that various algorithms come effectively and image data processing accurately from the obtainable detailed data of imaging system.Under the help of computing machine, the progress aspect Flame Image Process is normally carried out numeral or digitized image.
Be used to create the nuclear imaging of medical technology that the digital collection system comprises digital X-ray photography, computer x-ray tomography (CT) imaging, magnetic resonance imaging (MRI), ultrasonic (US) and for example positron emission x-ray tomography art (PET) and single photon radiates computed tomography (SPECT) of digital picture.Also can for example, analog image (for example typical X ray) produce digital picture according to analog image by being scanned into digital form.Yet, under the situation of not adding help, explain mass data in the digital picture for people, for example doctor normally difficulty with sesquipedalian.Computer-aided diagnosis (CAD) system help human, especially the medical science pathology visual, cut apart, play critical effect aspect detection, record and the report.
Produce digital picture according to representative and array of numerical digit values by the characteristic (for example gray-scale value or magnetic field intensity) of the related anatomical position spot correlation of specific array position.The set of described anatomical location points comprises image area.In 2D digital picture or sliced section, the array position of dispersion is known as pixel.Can make up three-dimensional digital image by the stacked sliced section of various constructing technology causes as known in the art.3D rendering is made up of discrete volume element (being also referred to as voxel), and described volume element is made of the pixel of 2D image.Can handle to determine various characteristics pixel or voxel properties about the patient anatomy relevant with this pixel or voxel.
In case make up and estimate to understand and cut zone and structure by analyzing pixel and/or voxel, just can improve the precision and the efficient of imaging system thus with subsequently the processing that utilizes region characteristic and feature and analytical applications in the relevant range.
One of more crucial CAD task comprises screening and the various types of cancers of early detection from volume data (for example, CT volume data).Many cancers, for example colon cancer are because the early detection of cancer knurl and removal have demonstrated the reduction of mortality ratio.Pathology is spherical typically or hemispheric aspect geometric configuration.In many cases, these spherical pathologies are attached on linearity or the piecewise linearity surface.Unfortunately, up to the late period of disease, the common detection of existing method does not go out the peculiar symptom of various cancers.The main target of the preventative cancer screening that therefore, shifts to an earlier date provides the early detection to peculiar symptom.
Summary of the invention
In one aspect of the invention, provide a kind of in the digital picture of colon identification polyp method, wherein said image comprise with 3 dimension spaces in the corresponding a plurality of intensity in voxel territory.Described method comprises: the image with one group 3 orthogonal axles is provided; At each voxel in the image, on first orientation, make cutting planes be centered in the center voxel with respect to image shaft; Determine the intersection of cutting planes and colon, and the track of check image in described intersection; And it is little and be to select in the intersection those voxels under the circular situation to be used for further analysis at described track.
In another aspect of the present invention, described track is with respect to little about two orders of magnitude of size of images.
In another aspect of the present invention, repeat describedly to center, determine and select step that at a plurality of cutting planes wherein each cutting planes is on the different orientations with respect to image shaft.
In another aspect of the present invention, described a plurality of cutting planes is evenly distributed on the orientation spheroid.
In another aspect of the present invention, select described cutting planes like this, make each plane have coordinate for (A, B, normal C), A wherein, B and C are the integers between-1 and 1, and A wherein, B and C can not be zero.
In another aspect of the present invention, A, B and C are subjected to | the restriction of A|+|B|+|C|＜=2.
In another aspect of the present invention, said method comprising the steps of: mark is selected for those voxels of further analysis, so that identification and mark one or more related bunch from selected voxel.
In another aspect of the present invention, described method comprises each the step of size in determining described one or more related bunch.
In another aspect of the present invention, described method comprises in described one or more related bunch each is made the step of the analysis of ring-type or spherical aspect.
Description of drawings
Fig. 1 describes the process flow diagram of the preferred method of the present invention;
Fig. 2 describes and the crossing cutting planes of colon;
Fig. 3 describes the axial slices of colon C T image;
Polyp in the colon C T image that Fig. 4 describes to discern by the use cutting planes;
Fig. 5 describes to be used to implement the exemplary computer system of the preferred embodiments of the invention.
Embodiment
Below illustrative examples of the present invention will be described.For the sake of clarity, actual all features of implementing are not described in this manual.Certainly it should be understood that, in the development of any this actual embodiment, must make the specific decision of many embodiments to realize developer's specific objective, for example conform to the constraint that relates to system and relate to commerce, described constraint will change according to the difference of implementation process.In addition, it should be understood that this development may be complicated with time-consuming, but will be the routine work of being engaged in for the those of ordinary skills that benefit from present disclosure.
Though the present invention allows various modifications and alternative form, shown particular of the present invention and it has been elaborated at this by the example in the accompanying drawing.Yet, it should be understood that, herein to the explanation of particular and be not intended to limit the invention to particular forms disclosed, but opposite, the present invention drops on covering as by all modifications in the appended spirit and scope defined in claim of the present invention, content of equal value and replacement scheme.
As used herein, term " image " refers to the multidimensional data that is made of discrete picture element (for example, the voxel of the pixel of 2D image and 3D rendering).For example, image can for example be the medical image by computer x-ray tomography, magnetic resonance imaging, the ultrasonic or object that any other medical image system well known by persons skilled in the art is gathered.Also can wait image is provided by the non-medical environment, such as for example remote detecting system, electron microscopy.Though can regard image as R
3To the function of R, but method of the present invention is not limited to this image, and method of the present invention can be applied to the image of any dimension, for example 2D picture or 3D volume.For 2 or 3 d image, image area is 2 or 3 dimension rectangular arrays typically, wherein can come each pixel of addressing or voxel with reference to one group 2 or 3 orthogonal axles.The present invention preferably carries out the upward execution of computer system (for example Pentium class personal computer) of the computer software of algorithm of the present invention in operation.Computing machine comprises processor, storer and various input-output apparatus.The digital picture of a series of representative chest volumes is inputed to computing machine.Term " numeral " and " digitizing " will refer to the numeral that obtains by the digital collection system or obtain by the conversion from analog image or the suitable image or the volume of digitized format as used herein.
Yet, before volume is cut, can carry out pre-service so that colon and other structural area in the image are separated to image.The high precision of algorithm is critical for the nodule detection of success, and pre-service reduces the complicacy of the domain of function of needs assessment usually.When pre-service based on just by the known features of the object of imaging the time, pre-service is normally more effective.Preprocessing Algorithm is well-known in the art, and for example comprises the technology of level and smooth, morphology and regularization.In the CT image, simple threshold value will be enough to distinguish tube chamber and tissue, but will need further pre-service to eliminate other border, for example extraneous air, lung, small intestine etc.
Referring now to Fig. 1, in a preferred embodiment of the invention, cutting planes is used to locate polyp in colon C T image.In step 101, at each voxel in the image, come volume is cut by the plane that has different orientation with respect to image shaft, wherein each plane all is centered on the voxel of being considered (being called the center voxel hereinafter).To the quantity of spendable orientation without limits, but in a preferred embodiment, have been found that one group 9 to 13 cutting planes on different orientations is just enough.Preferably, the orientation that is oriented in of these cutting planes should more or less be equally distributed on the spheroid.In a preferred embodiment, select described plane like this, make the normal on plane have coordinate (A, B, C), A wherein, B, C are the integers between-1 and 1, and are subjected to following restriction, promptly they can not all be zero.Exist and corresponding 13 planes of all possibilities, and 9 planes are corresponding to constraint | A|+|B|+|C|＜=2.Other standard that is used to select cutting planes also is feasible, and all within the scope of the invention.
In step 102, in each plane in these planes, check the track of volume then, this track is the value of this volume in this plane.Because image is pretreated probably so that colon and background area are separated, thus interested be the track that cutting planes and colon intersect.Little and circular track may be the part of polyp, because there is not other little circular configuration on colon wall.Have for each dimension in the image of about 512 voxels in typical sizes, little track is with little about two orders of magnitude, or about 3 voxels of each dimension.Therefore, in the sign that appears as polyp that in one group of cutting planes of a voxel, limits the track of little and circular regions.In the process of checking track, each voxel is considered just in time once in each plane.For every group of planar orientation, the plane of correct number is arranged definitely, make that each voxel in the neighborhood of center voxel is considered.The preferential selection of planar orientation is guaranteed to be in all voxels in the polyp and all is included in one of cutting planes of being centered in the center voxel.Under other non-integral planar orientation situation, voxel is in the polyp, but is not included in any one cutting planes.After having finished given plane, in step 103, will be labeled as the positive by those points in the little border circular areas that track limited with given orientation at each voxel.Therefore, for each planar orientation, each voxel all has picked chance as polyp.If 13 planar orientations are arranged, then will cut and wear each voxel, and each voxel has 13 chances and becomes the positive by 13 planes.At last, if found that on any orientation voxel is positive, then it is positive.It is all plane results' a scale-of-two " or ".After by each plane in this group cutting planes each voxel being cut, in step 104, abandon those points that maintenance is not labeled according to further analysis.
Each cutting planes at the different orientation in each voxel in the volume and this group cutting planes repeats to make the cutting planes of given orientation be centered in given center pixel, checks the track and the step of mark voxel to be used for further analyzing of the intersection of cutting planes and colon.
Fig. 2 illustrates the cutting planes that intersects with polyp.Track is the zone in the dotted line among this figure.Fig. 3 illustrates and presents by the axial slices indicated polyp of cross curve, original colon C T volume.
Fig. 4 illustrates the result of cutting planes detection algorithm, and wherein white point is the point that intersects in plane and the fraction of polyp on not being attached to colon wall.For this example, used 13 planar orientations that cover entire image, make each given voxel all be comprised in these 13 cutting planes.
In step 105, can use the mark process to find out voxel clusters in the image.Well-known process is to be associated to the branch mark (this procedure identification also marks each related composition in the image for connected component labeling, CCL) process.Usually, mark is the integer more than or equal to 1, and background is set to 0 simultaneously.The input of CCL algorithm is the output of binary picture, for example plane cutting process, wherein for example background be set to 0 and objects in voxel be set to 1.The output of CCL process is the image that has with the identical size of input picture, simultaneously will background voxels be set to 0 and make and have integer-valued object pixel since 1.Therefore, the CCL process is that each object in the image distributes a number.In case identified related bunch, just can be in step 106 they be further analyzed the feature of separating with the pattern that is identified for belonging to potential polyp and other pattern area, for example bunch size and ring-type or spherical. consist of the theme of the inventor's common pendent patent application for the spherical method of the structure of determining medical image: the U.S. Patent application No.10/915075 (agent numbers No.2003P12256US) " Method and System for Fast NormalizedCross-Correlation Between an Image and a Gaussian for DetectingSpherical Structures " that the U.S. Patent application No.10/915047 (agent numbers No.2003P12257US) " Method and System for Using StructureTensors to Detect Lung Nodules and Colon Polyps " that on August 10th, 2004 submitted to and on August 10th, 2004 submit to.
Should be appreciated that the present invention can make up with various forms of hardware, software, firmware, dedicated processes or its implements.In one embodiment, can with the invention process the application program that is comprised in effectively on the computer-readable program memory device with software.This application program can be uploaded on the machine that comprises any suitable framework and by this machine and carry out.
Referring now to Fig. 5,, be used to implement computer system 501 of the present invention and can especially comprise CPU (central processing unit) (CPU) 502, storer 503 and I/O (I/O) interface 504 according to one embodiment of the invention.Computer system 501 couples by I/O interface 504 and display 505 and various input equipment 506 (for example mouse and keyboard) usually.Support that circuit can comprise for example circuit of cache memory, power supply, clock circuit and communication bus.Storer 503 can comprise random-access memory (ram), ROM (read-only memory) (ROM), disc driver, tape drive etc. or its combination.Can be to be stored in the storer 503 and to carry out with the invention process with program 507 to handling from the signal of signal source 508 by CPU 502.Similarly, computer system 501 is general-purpose computing systems, and when carrying out program 507 of the present invention, it just becomes dedicated computer system.
It will also be appreciated that because the composition system unit described and some in the method step can be implemented with software in the accompanying drawing, so the actual connection between the system unit (or treatment step) can be according to the mode that the present invention is programmed and difference.If given instruction of the present invention mentioned herein, then the those of ordinary skill in the association area can be imagined these and similar embodiment or configuration of the present invention.
Above disclosed particular only be schematically because the present invention can be with different but revise and put into practice for benefiting from this those skilled in the art of instruction conspicuous equivalents.In addition, be not intended to be limited to shown here with different described in the following claim structures or the details of design.Therefore be apparent that and change or to revise top disclosed particular and think that all this variations are all in scope and spirit of the present invention.Therefore, in this protection of looking for described in claim below.
Claims (8)
- One kind in the digital picture of colon identification polyp method, wherein said image comprise with 3 dimension spaces in the corresponding a plurality of intensity in voxel territory, said method comprising the steps of:A plurality of cutting planes are provided, and each cutting planes all is in different orientations with respect to image shaft;At each voxel in the image, make each cutting planes in the described cutting planes be centered in the center voxel;At in described a plurality of cutting planes of each voxel in the image each, determine the intersection of cutting planes and colon, and the track of inspection cutting planes in described intersection; AndThe track of each cutting planes than the image in the cutting planes little two orders of magnitude and be under the circular situation, select those voxels in the intersection to be used for further analysis.
- 2. the method for claim 1, wherein 9 to 13 cutting planes are provided.
- 3. the method for claim 1, wherein described a plurality of cutting planes are evenly distributed on the orientation spheroid.
- 4. method as claimed in claim 3 wherein, is selected described cutting planes, make each plane have coordinate for (A, B, normal C), wherein A, B and C are the integers between-1 and 1, and wherein A, B and C can not be zero.
- 5. method as claimed in claim 4, wherein, A, B and C are subjected to | the restriction of A|+|B|+|C|＜=2.
- 6. the method for claim 1 also comprises step: mark is labeled and is used for those voxels of further analyzing, so that identification and mark one or more related bunch from the voxel of institute's mark.
- 7. method as claimed in claim 6 also comprises each the step of size in determining described one or more related bunch.
- 8. method as claimed in claim 6 also comprises in described one or more related bunch each is made the step of the analysis of ring-type or spherical aspect.
Applications Claiming Priority (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US50471403P | 2003-09-22 | 2003-09-22 | |
| US60/504,714 | 2003-09-22 | ||
| US10/945,130 | 2004-09-20 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1856802A CN1856802A (en) | 2006-11-01 |
| CN100481128C true CN100481128C (en) | 2009-04-22 |
Family
ID=37196011
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2004800272532A Expired - Fee Related CN100481128C (en) | 2003-09-22 | 2004-09-21 | Method and system for using cutting planes for colon polyp detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100481128C (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN100454340C (en) * | 2007-02-13 | 2009-01-21 | 上海交通大学 | Visual method for virtual incising tubular organ |
-
2004
- 2004-09-21 CN CNB2004800272532A patent/CN100481128C/en not_active Expired - Fee Related
Non-Patent Citations (4)
| Title |
|---|
| A Statistical 3-D Pattern Processing MethodforComputer-Aided Detection of Polyps in CT Colonography. GOKTURK S B ET AL.IEEE TRANSACTIONS ON MEDICAL IMAGING,Vol.vol.20 No.no.12. 2001 |
| A Statistical 3-D Pattern Processing MethodforComputer-Aided Detection of Polyps in CT Colonography. GOKTURK S B ET AL.IEEE TRANSACTIONS ON MEDICAL IMAGING,Vol.vol.20 No.no.12. 2001 * |
| Computer-aidedDiagnosisiSchemeforDetectionofPolypsatCTColonography. HIROYUKI YOSHIDA ET AL.Radiographics. online,Vol.vol.22 . 2002 |
| Computer-aidedDiagnosisiSchemeforDetectionofPolypsatCTColonography. HIROYUKI YOSHIDA ET AL.Radiographics. online,Vol.vol.22 . 2002 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1856802A (en) | 2006-11-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8031917B2 (en) | System and method for smart display of CAD markers | |
| Dey et al. | Soft computing based medical image analysis | |
| EP2443587B1 (en) | Systems for computer aided lung nodule detection in chest tomosynthesis imaging | |
| JP7005191B2 (en) | Image processing equipment, medical diagnostic imaging equipment, and programs | |
| CN111210897A (en) | Processing medical images | |
| JP2008521468A (en) | Digital medical image analysis | |
| JP2001511372A (en) | Method for improving accuracy in discriminative diagnosis in radiological examination | |
| CN1656499A (en) | Pulmonary nodule detection using cartwheel projection analysis | |
| JP2019082881A (en) | Image retrieval device, method and program | |
| US20220028510A1 (en) | Medical document creation apparatus, method, and program | |
| US8737699B2 (en) | Combinational computer aided diagnosis | |
| Yao et al. | Texture synthesis based thyroid nodule detection from medical ultrasound images: interpreting and suppressing the adversarial effect of in-place manual annotation | |
| CA2535404C (en) | Method and system for using cutting planes for colon polyp detection | |
| Kumar et al. | Fully automated segmentation of lung parenchyma using break and repair strategy | |
| US7397938B2 (en) | Method and system for fast normalized cross-correlation between an image and a Gaussian for detecting spherical structures | |
| CN1836258B (en) | Method and system for using structure tensors to detect lung nodules and colon polyps | |
| US20230005580A1 (en) | Document creation support apparatus, method, and program | |
| CN100481128C (en) | Method and system for using cutting planes for colon polyp detection | |
| CN1902640A (en) | System and method for filtering and automatic detection of candidate anatomical structures in medical images | |
| Seçkin et al. | Deep learning structures used in pulmonary cancer diagnosis | |
| US20230281810A1 (en) | Image display apparatus, method, and program | |
| US20240046028A1 (en) | Document creation support apparatus, document creation support method, and document creation support program | |
| US20230225681A1 (en) | Image display apparatus, method, and program | |
| US20230410305A1 (en) | Information management apparatus, method, and program and information processing apparatus, method, and program | |
| EP3794550B1 (en) | Comparison of a region of interest along a time series of images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20090422 Termination date: 20180921 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |