CN103236062B - Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse - Google Patents
Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse Download PDFInfo
- Publication number
- CN103236062B CN103236062B CN201310159783.3A CN201310159783A CN103236062B CN 103236062 B CN103236062 B CN 103236062B CN 201310159783 A CN201310159783 A CN 201310159783A CN 103236062 B CN103236062 B CN 103236062B
- Authority
- CN
- China
- Prior art keywords
- segmentation
- human brain
- blood vessel
- tumor
- magnetic resonance
- 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
- 230000011218 segmentation Effects 0.000 title claims abstract description 282
- 210000004204 blood vessel Anatomy 0.000 title claims abstract description 190
- 208000003174 Brain Neoplasms Diseases 0.000 title claims abstract description 127
- 230000005311 nuclear magnetism Effects 0.000 title claims abstract description 118
- 206010028980 Neoplasm Diseases 0.000 claims abstract description 245
- 210000004556 brain Anatomy 0.000 claims abstract description 105
- 238000000034 method Methods 0.000 claims abstract description 58
- 230000003902 lesion Effects 0.000 claims description 20
- 230000013011 mating Effects 0.000 claims description 11
- 238000009499 grossing Methods 0.000 claims description 7
- 238000003706 image smoothing Methods 0.000 claims description 7
- 230000002792 vascular Effects 0.000 claims description 7
- 230000009286 beneficial effect Effects 0.000 abstract description 4
- 238000002595 magnetic resonance imaging Methods 0.000 description 8
- 239000000284 extract Substances 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000003745 diagnosis Methods 0.000 description 5
- 230000008878 coupling Effects 0.000 description 4
- 238000010168 coupling process Methods 0.000 description 4
- 238000005859 coupling reaction Methods 0.000 description 4
- 238000001356 surgical procedure Methods 0.000 description 4
- 238000005452 bending Methods 0.000 description 3
- 239000008280 blood Substances 0.000 description 3
- 210000004369 blood Anatomy 0.000 description 3
- 230000002490 cerebral effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000007170 pathology Effects 0.000 description 3
- 230000002093 peripheral effect Effects 0.000 description 3
- 241001269238 Data Species 0.000 description 2
- 238000005481 NMR spectroscopy Methods 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 210000000988 bone and bone Anatomy 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 201000010099 disease Diseases 0.000 description 2
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000007917 intracranial administration Methods 0.000 description 2
- 210000000653 nervous system Anatomy 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 235000008733 Citrus aurantifolia Nutrition 0.000 description 1
- 235000011941 Tilia x europaea Nutrition 0.000 description 1
- 210000000577 adipose tissue Anatomy 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000011888 autopsy Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 208000012191 childhood neoplasm Diseases 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 239000004571 lime Substances 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 210000004072 lung Anatomy 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 210000000272 myelencephalon Anatomy 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 210000004872 soft tissue Anatomy 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 210000001519 tissue Anatomy 0.000 description 1
- 210000003462 vein Anatomy 0.000 description 1
- 210000001835 viscera Anatomy 0.000 description 1
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
Abstract
The invention discloses a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse, on the basis of human brain magnetic resonance tumor segmentation image, data according to the tumor region on above-mentioned human brain magnetic resonance tumor segmentation image obtain the experience segmentation threshold carrying out blood vessel segmentation process from human brain tumour's nuclear-magnetism storehouse, by analyzing human brain magnetic resonance tumor segmentation image and obtain blood vessel gray threshold according to above-mentioned experience segmentation threshold, blood vessel gray threshold is adopted to carry out blood vessel segmentation to human brain magnetic resonance tumor segmentation image.Above-mentioned segmenting system achieves the auto Segmentation of human brain tumour's magnetic resonance image (MRI) blood vessel, avoid artificial participation, improve the accuracy of segmentation, and the above-mentioned method comparison split the blood vessel of human brain magnetic resonance tumor segmentation image simply, is more beneficial to the commercialization of system.
Description
Technical field
The present invention relates to Iamge Segmentation field, particularly relate to a kind of magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse and a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse.
Background technology
Brain tumor is one of disease common in nervous system, has very large harm to the function of human nervous system.In recent years, the ICT (intracranial tumor) incidence of disease is in rising trend.According to statistics, ICT (intracranial tumor) accounts for 5% of general tumour, account for 70% of pediatric tumor, the great threat life security of people.When comprising blood vessel around tumour, tumour can cause the pathology of blood vessel, and simultaneously in order to provide blood to tumour, the distribution of tumour peripheral vessels can increase.In cerebral surgery operation, blood vessel must preferentially be avoided in operation, particularly sustainer, cardinal vein blood vessel.Because knub position possibility is very many, the situation of tumour parcel blood vessel is very general, and therefore how to allow doctor understand the relative position of tumour and blood vessel intuitively, the success or not for cerebral surgery operation is most important.
Magnetic resonance image (MRI) is at present for one of best mode confirming knub position, on the basis of magnetic resonance image (MRI), can by the principle of graphical analysis, the position of blood vessel is divided in magnetic resonance image (MRI), again in conjunction with technology such as lesion segmentation, 3D Model Reconstruction, can rebuild human brain, tumour and blood vessel on 3-D view, help the relative position of doctor to tumour and blood vessel to get information about, carry out for what perform the operation the effect playing assistance.
Image segmentation algorithm is widely used in magnetic resonance image analysis field, and especially, magnetic resonance image (MRI) is gray level image, and the inner various institutional framework of human brain is intricate, and its magnetic resonance image (MRI) lines detailed information formed is more, and segmentation has certain difficulty.Reach good segmentation effect, cutting procedure needs strict defining, otherwise simply calculates according to gray scale, may occur that noise is many, split not accurate enough problem.
Current blood vessel segmentation technology is generally divided into three major types: based on profile, based on region with based on particular theory and particular tool.Carry out blood vessel segmentation based on region-growing method, the basic mode adopting artificial neural network, operator needs the mode providing Seed Points and Luminance edge threshold value to carry out semi-automatic segmentation.Adopt the mode of artificial intelligence and vascular pattern to carry out blood vessel segmentation, calculate accuracy better, but the complexity of algorithm is quite high.
But in actual applications, the scanning number of plies of magnetic resonance image (MRI) reaches tens, layer up to a hundred, if require each aspect needs operator manually to participate in, operation is trouble quite, simultaneously very high to the requirement of operator's professional knowledge, and has suitable subjectivity; Adopt the dividing method that algorithm complex is high, although it is better to calculate accuracy, algorithm complex is too high is unfavorable for commercialization.
Summary of the invention
Based on this, to be necessary for the dividing method troublesome poeration needing operator manually to participate in tumour magnetic resonance image (MRI) blood vessel segmentation process and very high and there is suitable subjectivity and the dividing method taking algorithm complex high, the too high problem being unfavorable for commercialization of algorithm complex to the requirement of operator, to provide a kind of algorithm complex low and the magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse not needing operator manually to participate in.
Meanwhile, a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse is also provided.
Based on the magnetic resonance image (MRI) blood vessel segmentation method in human brain tumour's nuclear-magnetism storehouse, comprise the steps:
Mark to obtain human brain magnetic resonance tumor segmentation image to the tumor region in initial human brain magnetic resonance image (MRI);
According to the characteristic matching human brain tumour nuclear-magnetism storehouse of described tumor region, obtain the experience segmentation threshold described human brain magnetic resonance tumor segmentation image being carried out to blood vessel segmentation process from described human brain tumour's nuclear-magnetism storehouse;
Described human brain magnetic resonance tumor segmentation image is analyzed and obtained blood vessel gray threshold according to described experience segmentation threshold, and adopts described blood vessel gray threshold to carry out blood vessel segmentation to described human brain magnetic resonance tumor segmentation image.
Wherein in an embodiment, described intensity profile human brain magnetic resonance tumor segmentation image being analyzed to the non-tumor region specifically obtaining human brain magnetic resonance tumor segmentation image, and adopting gray probability function f (m) to represent described intensity profile, gray probability function f (m) represents that gray shade scale is the ratio of quantity shared by described non-tumor region of the pixel of m; Described experience segmentation threshold Th represent non-vascular region in described non-tumor region account for described non-tumor region ratio and according to following formulae discovery blood vessel gray threshold:
Wherein, 1≤m≤n, when calculating m=1 to m=n f (m) with threshold grey scale value when being greater than Th, and using described threshold grey scale value as described blood vessel gray threshold.
Wherein in an embodiment, also comprise the step setting up described human brain tumour's nuclear-magnetism storehouse, described human brain tumour's nuclear-magnetism storehouse comprises human brain tumour's area data, corresponding tumor's profiles data and corresponding experience segmentation threshold.
Wherein in an embodiment, the feature of described tumor region comprises tumor area and tumor's profiles, and described tumor area and tumor's profiles specifically mate with the human brain tumour's area data in described human brain tumour's nuclear-magnetism storehouse, corresponding tumor's profiles data by the described characteristic matching human brain tumour nuclear-magnetism storehouse according to tumor region respectively.
Wherein in an embodiment, described step of carrying out outline with human brain tumour's nuclear-magnetism storehouse is carry out outline by the outline method of Hu not bending moment.
Wherein in an embodiment, also comprise the step in the largest connected territory of the image after calculating described blood vessel segmentation.
Wherein in an embodiment, also comprise the step that the image after to the largest connected territory of described calculating is smoothing.
Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse, comprising:
Lesion segmentation module, obtains human brain magnetic resonance tumor segmentation image for marking the tumor region in initial human brain magnetic resonance image (MRI);
Human brain tumour's nuclear-magnetism library module, for storing and providing human brain magnetic resonance tumor image and related data;
Human brain tumour's nuclear-magnetism storehouse matching module, connect described lesion segmentation module and described human brain tumour's nuclear-magnetism library module respectively, carry out mating obtaining with the related data in described human brain tumour's nuclear-magnetism library module for the related data of the tumor region by described human brain magnetic resonance tumor segmentation image and the experience segmentation threshold that image carries out blood vessel segmentation is split to described human brain magnetic resonance tumor;
Blood vessel segmentation module, connect described lesion segmentation module and described human brain tumour's nuclear-magnetism storehouse matching module respectively, for analyzing described human brain magnetic resonance tumor segmentation image and obtain blood vessel gray threshold according to described experience segmentation threshold, and described blood vessel gray threshold is adopted to carry out blood vessel segmentation to described human brain magnetic resonance tumor segmentation image.
Wherein in an embodiment, described human brain tumour's nuclear-magnetism storehouse matching module comprises further:
Tumor area matching unit, connects described lesion segmentation module and described human brain tumour's nuclear-magnetism library module respectively, for extracting the area of the tumor region of described human brain magnetic resonance tumor segmentation image and carrying out area matched with described human brain tumour's nuclear-magnetism library module;
Tumor's profiles matching unit, connect described lesion segmentation module, described human brain tumour's nuclear-magnetism library module respectively, for extracting the outline data of the tumor region of described human brain magnetic resonance tumor segmentation image and carrying out outline with described human brain tumour's nuclear-magnetism library module;
The match is successful judging unit, connect described tumor area matching unit, described tumor's profiles matching unit and described blood vessel segmentation module respectively, matched data for being obtained by described tumor area matching unit and described tumor's profiles matching unit judges whether that the match is successful, and sends to described blood vessel segmentation module by what obtain after the match is successful to the experience segmentation threshold that described human brain magnetic resonance tumor segmentation image carries out blood vessel segmentation process.
Wherein in an embodiment, described blood vessel segmentation module comprises further:
Blood vessel gray threshold computing unit, described blood vessel gray threshold computing unit connects described lesion segmentation module and described human brain tumour's nuclear-magnetism storehouse matching module respectively, for analyzing described human brain magnetic resonance tumor segmentation image and obtain described blood vessel gray threshold according to described experience segmentation threshold;
Vessel segmentation unit, connects described blood vessel gray threshold computing unit, carries out blood vessel segmentation for adopting described blood vessel gray threshold to described human brain magnetic resonance tumor segmentation image.
Wherein in an embodiment, also comprise largest connected territory computing module, described largest connected territory computing module connects described blood vessel segmentation module, for calculating the largest connected territory of the image after described blood vessel segmentation.
Wherein in an embodiment, also comprise image smoothing module, largest connected territory computing module described in described image smoothing model calling, for smoothing to the image behind the largest connected territory of described calculating.
The above-mentioned magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse and the magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse, on the basis of human brain magnetic resonance tumor segmentation image, data according to the tumor region on above-mentioned human brain magnetic resonance tumor segmentation image obtain experience segmentation threshold human brain magnetic resonance tumor segmentation image being carried out to blood vessel segmentation process from human brain tumour's nuclear-magnetism storehouse, by analyzing human brain magnetic resonance tumor segmentation image and obtain blood vessel gray threshold according to above-mentioned experience segmentation threshold, above-mentioned blood vessel gray threshold is adopted to carry out blood vessel segmentation to human brain magnetic resonance tumor segmentation image.By the above-mentioned magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse and the auto Segmentation achieving human brain tumour's magnetic resonance image (MRI) blood vessel based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse, avoid artificial participation, improve the accuracy of segmentation, and the above-mentioned method comparison split the blood vessel of human brain magnetic resonance tumor segmentation image simply, is more beneficial to the commercialization of system.
Accompanying drawing explanation
Fig. 1 is the magnetic resonance image (MRI) blood vessel segmentation method flow diagram based on human brain tumour's nuclear-magnetism storehouse of one embodiment of the invention;
Fig. 2 is the process flow diagram of step S130 embodiment illustrated in fig. 1;
Fig. 3 is step S130 embodiment illustrated in fig. 1 process flow diagram in another embodiment;
Fig. 4 is the magnetic resonance image (MRI) blood vessel segmentation system module figure based on human brain tumour's nuclear-magnetism storehouse of one embodiment of the invention;
Fig. 5 is the magnetic resonance image (MRI) blood vessel segmentation system module figure based on human brain tumour's nuclear-magnetism storehouse of another embodiment of the present invention.
Embodiment
A kind of magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse and a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse, by setting up human brain tumour's nuclear-magnetism storehouse, when carrying out blood vessel segmentation by mating with the data in above-mentioned human brain tumour's nuclear-magnetism storehouse, find experience segmentation threshold when blood vessel segmentation carried out to human brain magnetic resonance tumor segmentation image and obtain by carrying out analyzing and processing to the non-tumor region of above-mentioned human brain magnetic resonance tumor segmentation image blood vessel gray threshold when image carries out blood vessel segmentation is split to above-mentioned human brain magnetic resonance tumor, blood vessel segmentation is carried out to human brain tumour's magnetic resonance image (MRI), foundation when carrying out cerebral surgery operation as doctor.By carrying out blood vessel segmentation based on the magnetic resonance image (MRI) blood vessel segmentation method in human brain tumour's nuclear-magnetism storehouse and a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse, avoiding the artificial participation of the operator when carrying out blood vessel segmentation, improve the accuracy of blood vessel segmentation, the algorithm complex of above-mentioned partitioning algorithm is lower, is conducive to the commercialization of system.
Below in conjunction with drawings and Examples, a kind of magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse of the present invention and a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse are described in more detail.
Shown in Fig. 1, it is the magnetic resonance image (MRI) blood vessel segmentation method flow diagram based on human brain tumour's nuclear-magnetism storehouse of one embodiment of the invention.
With reference to figure 1, a kind of magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse, for carrying out blood vessel segmentation to human brain magnetic resonance tumor segmentation image, specifically comprises the steps:
Step S110: set up human brain tumour's nuclear-magnetism storehouse.
The 3 D anatomical image that magnetic resonance imaging (MagneticResonanceImaging, MRI) can provide Human autopsy tissues clear and intuitive, has higher soft tissue resolution, is the topmost image check means of current brain tumor diagnosis.
Along with the application of magnetic resonance imaging in brain tumor diagnosis is increasing, use magnetic resonance imaging brain tumor is studied, Diagnosis and Treat time can produce a large amount of data, comprising magnetic resonance imaging produce view data and doctor undertaken operating and diagnosing by magnetic resonance image (MRI) time produce some related datas.By setting up human brain tumour's nuclear-magnetism storehouse and the related data of some magnetic resonance imagings of doctor in brain tumor diagnosis process and corresponding tumor image data are added up by human brain tumour's database, to apply in follow-up study with diagnosis.
The data in above-mentioned human brain tumour's nuclear-magnetism storehouse comprise the area data of tumour in human brain magnetic resonance image (MRI), the outline data of corresponding tumour and to corresponding experience segmentation threshold lesion segmentation image being carried out to blood vessel segmentation.The blood vessel gray threshold of acquisition when above-mentioned experience segmentation threshold refers to that doctor carries out blood vessel segmentation by human brain magnetic resonance image (MRI).When adding up above-mentioned related data, by the experience segmentation threshold one_to_one corresponding of the area data of the tumour of extraction, outline data and correspondence.Reference table 1, the tumour data of such as numbering 00012, corresponding size is 54 × 10
2individual pixel, outline data corresponds to SHAPE00012, and corresponding blood vessel segmentation empirical value is 0.6.When using above-mentioned human brain nuclear-magnetism storehouse, by mating tumor area data and tumor's profiles data respectively, thus obtain corresponding experience segmentation threshold.
In other examples, the experience segmentation threshold that blood vessel is split is represented by other representations.
In other examples, above-mentioned human brain tumour's nuclear-magnetism storehouse also can comprise other data characteristicses relevant to tumour magnetic resonance image (MRI).
Table 1 human brain tumour nuclear-magnetism database data signal table
| Numbering (ID) | Area (pixel set)/10 2 | Outline data (point set) | Blood vessel segmentation empirical value (0-1) |
| 00011 | 60 | SHAPE00011 | 0.3 |
| 00012 | 54 | SHAPE00012 | 0.6 |
| 00013 | 80 | SHAPE00013 | 0.5 |
| 00014 | 102 | SHAPE00014 | 0.6 |
| …… | …… | …… | …… |
Step S120: mark to obtain human brain magnetic resonance tumor segmentation image to the tumor region in initial human brain magnetic resonance image (MRI).
Above-mentioned human brain magnetic resonance tumor segmentation image refers to and uses a kind of number of greyscale levels not identical with this figure gray scale to show the tumor region in image, conveniently extract the relevant information of corresponding tumor region and non-tumor region, comprise the intensity profile etc. of the data such as area or profile of tumor region, non-tumor region.
Concrete, the tumor region in initial human brain magnetic resonance image (MRI) is marked and completes by Iamge Segmentation.Above-mentioned Iamge Segmentation utilizes the imaging grey value difference of tumor region and non-tumor region in magnetic resonance image (MRI), is split by tumor region by arranging intensity slicing threshold value.
Step S130: according to the characteristic matching human brain tumour nuclear-magnetism storehouse of tumor region, obtains experience segmentation threshold human brain magnetic resonance tumor segmentation image being carried out to blood vessel segmentation process from human brain tumour's nuclear-magnetism storehouse.
Concrete, the feature of above-mentioned tumor region comprises the area data of tumor region, outline data etc.In order to realize the auto Segmentation of blood vessel, first algorithm will be automatically found the blood vessel gray threshold split blood vessel.During owing to comprising blood vessel around tumour, tumour can cause the pathology of blood vessel, and simultaneously in order to provide blood to tumour, the distribution of tumour peripheral vessels can increase.The probability that increases based on blood vessel follows tumour at the clinical inference that nuclear-magnetism image upper section amasss and contour feature is directly proportional, by carrying out mating with the tumour data in human brain tumour's nuclear-magnetism storehouse the experience segmentation threshold obtained when splitting blood vessel.
Concrete, mate with the related data in human brain tumour's nuclear-magnetism storehouse by tumour sectional area and tumor's profiles data, in regulation matching domain, find the most close tumor image, get experience segmentation threshold corresponding in such cases.
Shown in Fig. 2, it is the process flow diagram of step S130 shown in Fig. 1.
Concrete, with reference to figure 2, the step obtaining the default threshold value of blood vessel segmentation of human brain magnetic resonance tumor segmentation image above by coupling human brain tumour's nuclear-magnetism storehouse specifically comprises the steps:
Step S131: the area data and the outline data that extract human brain magnetic resonance tumor segmentation image tumor region respectively.
Concrete, the area of the tumor region of above-mentioned human brain magnetic resonance tumor segmentation image is obtained by image processing tool such as Matlab.Concrete, above-mentioned area can be expressed as the quantity of tumor region pixel.In other examples, above-mentioned area also can represent with other method for expressing.
Concrete, above-mentioned outline data represents the point set of tumor region contour edge.The outline data that straight-line detection, Ha Er (Haar) wavelet conversion and Tuscany (Canny) rim detection scheduling algorithm extract above-mentioned tumor area is converted by the contours extract algorithm of mathematical morphology, Hough (Hough).Concrete, obtain corresponding tumor's profiles data by image processing tool such as Matlab, CorelDRAW.
Step S132: mate with the related data in human brain tumour's nuclear-magnetism storehouse respectively according to the area data of tumor region extracted and outline data.
Area data according to the tumor region extracted is mated with the related data in human brain tumour's nuclear-magnetism storehouse, and different area matched methods can be selected to mate.Concrete, mate by the method obtaining area difference, then by judging area difference whether in default difference in areas threshold range.If the area difference obtained is in above-mentioned default difference in areas threshold range, the match is successful to represent area, otherwise area matched unsuccessful.
Outline data according to the tumor region extracted mates with the related data in human brain tumour's nuclear-magnetism storehouse, and different outline methods can be selected to mate.Concrete, by obtaining matching value based on the outline method of Hu not bending moment.According to the matching value obtained, judge it whether in the outline threshold value preset, if, represent outline success, otherwise outline is unsuccessful.
Step S133: judge that whether coupling is successful.
Mated respectively by tumor area data and outline data, if the match is successful respectively for above-mentioned tumor area data and outline data, and in the above-mentioned respectively data group that the match is successful, there is the data group of at least one group of overlap, then represent that the match is successful for the above-mentioned human brain magnetic resonance tumor segmentation tumor region of image and human brain tumour's nuclear-magnetism storehouse, otherwise mate unsuccessful.If mate unsuccessful, then perform step S134, if the match is successful, then perform step S135.
Step S134: manual blood vessel segmentation is carried out to above-mentioned human brain magnetic resonance tumor segmentation image.
As long as above-mentioned tumor area data and outline data respectively the match is successful, then represent that the tumor region of above-mentioned human brain magnetic resonance tumor segmentation image and human brain tumour's nuclear-magnetism storehouse do not have that the match is successful.When mating with the data in corresponding tumour nuclear magnetic data storehouse, if do not match similar data, then manually determine blood vessel segmentation threshold value, and by being updated to after relevant data screening in corresponding human brain tumor nuclear-magnetism storehouse, then terminate this matching operation.
Step S135: obtain experience segmentation threshold human brain magnetic resonance tumor segmentation image being carried out to blood vessel segmentation process.If the tumor region of above-mentioned human brain magnetic resonance tumor segmentation image the match is successful with human brain tumour's nuclear-magnetism storehouse, experience segmentation threshold accordingly in the data group obtaining above-mentioned overlap, then terminates this matching operation.
In other examples, the tumor region of above-mentioned human brain magnetic resonance tumor segmentation image and human brain tumour's nuclear-magnetism storehouse may exist when the match is successful and comprise more than one group of data group that the match is successful, one group that now selects matching rate higher, as one group of data group that the match is successful, is therefrom read corresponding experience segmentation threshold.
Shown in Fig. 3, it is step S130 embodiment illustrated in fig. 1 process flow diagram in another embodiment.
With reference to figure 3, carry out in the process of mating at the tumor region of human brain magnetic resonance tumor segmentation image with human brain tumour's nuclear-magnetism storehouse, also can be mated by other matching process.Concrete, comprise the steps:
Step S1301: the area extracting the tumor region of human brain magnetic resonance tumor segmentation image.Concrete, the area of the tumor region of above-mentioned human brain magnetic resonance tumor segmentation image is obtained by image processing tool such as Matlab.Concrete, above-mentioned area can be expressed as the quantity of tumor region pixel.
Step S1302: carry out area matched with human brain tumour's nuclear-magnetism storehouse.Concrete, mate by the method obtaining area difference.
Step S1303: judge area matched whether successful.Concrete, in above-mentioned area matched process, namely area difference represents area in 50 pixels, and the match is successful, otherwise mate unsuccessful.
In other examples, above-mentioned area difference according to circumstances can carry out the setting of different difference.Such as, if by using other modes outside pixel number quantity set to represent area, represented the area of above-mentioned tumour by a threshold value, then its difference in areas selects corresponding threshold difference to represent.
If not, then perform step S1304, if so, then perform step S1305.
Step S1304: manual blood vessel segmentation is carried out to above-mentioned human brain magnetic resonance tumor segmentation image.
If area matched unsuccessful, to represent in human brain tumour's nuclear-magnetism storehouse the data do not split image tumor region area with above-mentioned human brain magnetic resonance tumor and match.If do not match similar data, then manually determine blood vessel segmentation threshold value, and by being updated to after relevant data screening in corresponding human brain tumor nuclear-magnetism storehouse, then terminate this matching operation.
Step S1305: if area matched success, then extract the outline data of tumor region.
Step S1306: carry out outline with human brain tumour's nuclear-magnetism storehouse.If area matched success, whether the outline data of the tumour extracted corresponding to the area data tumour that the match is successful further mates.
Step S1307: judge that whether outline is successful.Concrete, in the process of above-mentioned outline, by obtaining matching value based on the outline method of Hu not bending moment.
In other examples, also can be mated the tumor's profiles in above-mentioned human brain magnetic resonance tumor segmentation image and the tumor's profiles data in human brain tumour's nuclear-magnetism storehouse by different outline methods, obtain corresponding matching value.
If not, then exit this matching operation, if so, perform step S1308.
Step S1308: if outline success, then obtain the blood vessel segmentation default threshold value of corresponding blood vessel segmentation empirical value as human brain magnetic resonance tumor segmentation image.
By mating with the data in human brain tumour's nuclear-magnetism storehouse, first the coupling of tumor area is carried out, the coupling of the area data tumor's profiles feature that the match is successful is carried out again further if the match is successful, if the match is successful further for tumor's profiles, obtain the default threshold value of blood vessel segmentation that corresponding blood vessel segmentation empirical value is split as human brain magnetic resonance tumor, then exit this matching operation.
Further, when carrying out area matched, if meet the data not only a group in human brain tumour's nuclear-magnetism storehouse of above-mentioned area matched successful ranges, then one group of the highest data of matching rate are selected to continue next step outline.In other examples, also matching rate is higher in matching range several groups of data can be selected to enter next step outline, avoid occurring that the outline data phenomenon that the match is successful within the scope of corresponding outline of the data of the correspondence group that one group of data that area matched rate is the highest unsuccessful and area matched rate of outline when carrying out outline is time high occurs.
In other examples, also corresponding blood vessel segmentation empirical value can be obtained by carrying out the area matched method of tumor region volume again after the outline of first to carry out tumor region.
In other examples, different matching process and match-on criterion can be used, be not limited only to above-mentioned area matchedly with the method for outline, the tumour data that above-mentioned human brain magnetic resonance tumor is split in image to be mated with the data in human brain tumour's nuclear-magnetism storehouse, the correlation values needed for acquisition.
Step S140: analysiss is carried out to human brain magnetic resonance tumor segmentation image and empirically segmentation threshold obtains blood vessel gray threshold, and adopt blood vessel gray threshold to split image to human brain magnetic resonance tumor to carry out blood vessel segmentation.
Concrete, above-mentioned intensity profile human brain magnetic resonance tumor segmentation image being analyzed to the non-tumor region referring to acquisition human brain magnetic resonance tumor segmentation image, and adopt gray probability function f (m) to represent above-mentioned intensity profile, gray probability function f (m) represents that gray shade scale is the ratio shared in above-mentioned non-tumor region of the quantity of the pixel of m, and above-mentioned experience segmentation threshold Th represents ratio in above-mentioned non-tumor region shared by non-vascular region and according to following formulae discovery blood vessel gray threshold:
Wherein, 1≤m≤n, when calculating m=1 to m=n f (m) with threshold grey scale value when being greater than Th, and using above-mentioned threshold grey scale value as blood vessel gray threshold.
Carry out blood vessel segmentation according to above-mentioned blood vessel gray threshold to above-mentioned human brain magnetic resonance tumor segmentation image, during segmentation, at the non-tumor region of above-mentioned human brain magnetic resonance tumor segmentation image, the pixel that number of greyscale levels is greater than above-mentioned blood vessel gray threshold is blood vessel.
According to the principle of magnetic resonance imaging, nuclear magnetic resonance image GTG feature is, magnetic resonance signal is stronger, then brightness is larger, and the signal of magnetic resonance is weak, then brightness is also little; Brightness shows on image from big to small successively from white, grey to black.Various nuclear magnetic resonance image GTG feature of organizing is as follows: adipose tissue, cancellous bone are in white; Myelencephalon, marrow are lime look; Internal organ, muscle are pale liquid; Normal speed stream blood is black; Cortex of bone, gas, gassiness lung are black.Generally, during magnetic resonance imaging, the gray scale of blood vessel is higher, and its brightness is comparatively large, is presented on that image to be expressed as blood vessel brighter.
Step S150: the largest connected territory calculating the image after blood vessel segmentation.Above-mentioned connected domain refers to the angiosomes after segmentation and non-vascular zonule noise, and the largest connected territory of the image after above-mentioned calculating blood vessel segmentation is the non-vascular zonule noise produced when reducing segmentation, finds out the angiosomes after segmentation.
Step S160: smoothing to the image calculated behind largest connected territory.Smoothing to the image calculated behind largest connected territory, export the result of last blood vessel segmentation.
In the acquisition and transmitting procedure of image, original image can be subject to the interference of various noise, makes image quality decrease.In order to restraint speckle, improve picture quality, to the smoothing process of image.The process suppressed or eliminate these noises and improve picture quality is called the level and smooth of image.The method of carrying out image smoothing mainly comprises spatial domain low-pass filtering, frequency domain low-pass ripple method etc.
Blood vessel segmentation is carried out by the above-mentioned magnetic resonance image (MRI) blood vessel segmentation method based on human brain tumour's nuclear-magnetism storehouse, the algorithm used is relatively simple, be beneficial to commercialization, by carrying out mating with the data in human brain tumour's nuclear-magnetism storehouse the artificial participation decreasing operator in operating process, improve the accuracy of operation.
In other examples, for except brain tumor, the tumor disease that can impact tumour peripheral vessels density equally, carry out needing equally in surgical procedures by blood vessel segmentation out for operation in avoid vessel position time, also relevant tumour database can be set up, some related datas when these tumours being carried out to resonance manipulation are carried out sorting-out in statistics, can use when carrying out blood vessel segmentation and obtain the default threshold value of corresponding blood vessel segmentation above by carrying out mating to related neoplasms database the method obtaining corresponding blood vessel segmentation threshold value.
Shown in Fig. 4, be the magnetic resonance image (MRI) blood vessel segmentation system module figure based on human brain tumour's nuclear-magnetism storehouse of one embodiment of the invention.
With reference to figure 4, a kind of magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse, comprises lesion segmentation module 110, human brain tumour's nuclear-magnetism library module 120, human brain tumour's nuclear-magnetism storehouse matching module 130 and blood vessel segmentation module 140.
Above-mentioned human brain tumour's nuclear-magnetism storehouse matching module 130 connects the signal output part of lesion segmentation module 110, human brain tumour's nuclear-magnetism library module 120 respectively, and blood vessel segmentation module 140 connects the signal output part of lesion segmentation module 110 and human brain tumour's nuclear-magnetism storehouse matching module 130 respectively.
Lesion segmentation mould 110 marks to obtain human brain magnetic resonance tumor segmentation image to the tumor region in initial human brain magnetic resonance image (MRI), above-mentioned human brain tumour's nuclear-magnetism library module 120 is for storing and providing human brain magnetic resonance tumor image and related data, human brain tumour's nuclear-magnetism storehouse matching module 130 carries out mating obtaining with the related data in human brain tumour's nuclear-magnetism library module 120 by the related data of tumor region of human brain magnetic resonance tumor segmentation image splits to human brain magnetic resonance tumor the experience segmentation threshold that image carries out blood vessel segmentation, blood vessel segmentation module 140 pairs of human brain magnetic resonance tumor segmentation images are analyzed and obtain blood vessel gray threshold according to above-mentioned experience segmentation threshold, and adopt above-mentioned blood vessel gray threshold to carry out blood vessel segmentation to human brain magnetic resonance tumor segmentation image.
The above-mentioned magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse is on the basis of human brain magnetic resonance tumor segmentation image, data according to the tumor region on above-mentioned human brain magnetic resonance tumor segmentation image obtain experience segmentation threshold human brain magnetic resonance tumor segmentation image being carried out to blood vessel segmentation process from human brain tumour's nuclear-magnetism storehouse, by analyzing human brain magnetic resonance tumor segmentation image and obtain blood vessel gray threshold according to above-mentioned experience segmentation threshold, above-mentioned blood vessel gray threshold is adopted to carry out blood vessel segmentation to human brain magnetic resonance tumor segmentation image.The above-mentioned magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse achieves the auto Segmentation of human brain tumour's magnetic resonance image (MRI) blood vessel, avoids the artificial participation of operator, improves the accuracy of segmentation; And the method comparison of above-mentioned blood vessel segmentation is simple, is more beneficial to the commercialization of system.
Concrete, above-mentioned human brain tumour's nuclear-magnetism library module 120 comprises the area data of tumour in human brain magnetic resonance image (MRI), corresponding tumor's profiles data and corresponding blood vessel segmentation empirical value.In other examples, for needing the related neoplasms pathology confirming tumour periphery vessel position equally when performing the operation, relevant tumour nuclear-magnetism storehouse can be set up.Carrying out operation consent, using relevant tumour nuclear-magnetism storehouse to carry out Data Matching and obtain the default threshold value of blood vessel segmentation.
Shown in Fig. 5, be the magnetic resonance image (MRI) blood vessel segmentation system module figure based on human brain tumour's nuclear-magnetism storehouse of another embodiment of the present invention.
Concrete, with reference to figure 5, above-mentioned human brain tumour's nuclear-magnetism storehouse matching module 130 comprises tumor area matching unit 132, tumor's profiles matching unit 134 and the match is successful judging unit 136.Above-mentioned tumor area matching unit 132 connects lesion segmentation module 110 and human brain tumour's nuclear-magnetism library module 120, tumor's profiles matching unit 134 connects lesion segmentation module 110 and human brain tumour's nuclear-magnetism library module 120 respectively, and the match is successful, and judging unit 136 connects tumor area matching unit 132, tumor's profiles matching unit 134 and blood vessel segmentation module 136 respectively.
Above-mentioned tumor area matching unit 132 extracts the area of the tumor region of human brain magnetic resonance tumor segmentation image and carries out area matched with human brain tumour's nuclear-magnetism library module 120, tumor's profiles matching unit 134 extracts the outline data of the tumor region of human brain magnetic resonance tumor segmentation image and carries out outline with human brain tumour's nuclear-magnetism library module 120, the match is successful, and matched data that judging unit 136 obtained by tumor area matching unit 132 and tumor's profiles matching unit 134 judges whether that the match is successful, and by what obtain after the match is successful, blood vessel segmentation module 140 is sent to the experience segmentation threshold that human brain magnetic resonance tumor segmentation image carries out blood vessel segmentation process.
In other examples, also can by first to carry out area matched carry out outline again mode or characteristic that advanced row outline is carrying out the tumor region that area matched mode completes in human brain magnetic resonance tumor segmentation image mate with the related data in human brain tumour's nuclear-magnetism library module 120.
Concrete, with reference to figure 5, above-mentioned blood vessel segmentation module 140 comprises blood vessel gray threshold computing unit 142 and vessel segmentation unit 144.Blood vessel gray threshold computing unit 142 pairs of human brain magnetic resonance tumor segmentation images are analyzed and obtain blood vessel gray threshold according to above-mentioned experience segmentation threshold, and vessel segmentation unit 144 adopts above-mentioned blood vessel gray threshold to carry out blood vessel segmentation to above-mentioned human brain magnetic resonance tumor segmentation image.
Concrete, above-mentioned blood vessel gray threshold computing unit 142 pairs of human brain magnetic resonance tumor segmentation images analyze the intensity profile referring to the non-tumor region obtaining human brain magnetic resonance tumor segmentation image, and adopt gray probability function f (m) to represent above-mentioned intensity profile, gray probability function f (m) represents that gray shade scale is the ratio shared in above-mentioned non-tumor region of the quantity of the pixel of m, and above-mentioned experience segmentation threshold Th represents ratio in above-mentioned non-tumor region shared by non-vascular region and according to following formulae discovery blood vessel gray threshold:
Wherein, 1≤m≤n, when calculating m=1 to m=n f (m) with threshold grey scale value when being greater than Th, and using above-mentioned threshold grey scale value as blood vessel gray threshold.
Carry out blood vessel segmentation according to above-mentioned blood vessel gray threshold to above-mentioned human brain magnetic resonance tumor segmentation image, during segmentation, at the non-tumor region of above-mentioned human brain magnetic resonance tumor segmentation image, the pixel that number of greyscale levels is greater than above-mentioned blood vessel gray threshold is blood vessel.
Further, with reference to figure 5, the above-mentioned magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse comprises largest connected territory computing module 150 further, above-mentioned largest connected territory computing module 150 connects above-mentioned blood vessel segmentation module 140, for calculating the largest connected territory of the image after above-mentioned blood vessel segmentation.
Further, with reference to figure 5, the above-mentioned magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse comprises image smoothing module 160 further, above-mentioned image smoothing module 160 connects above-mentioned largest connected territory computing module 150, for smoothing to the image behind the largest connected territory of above-mentioned calculating.
The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (5)
1., based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse, it is characterized in that, comprising:
Lesion segmentation module, obtains human brain magnetic resonance tumor segmentation image for marking the tumor region in initial human brain magnetic resonance image (MRI);
Human brain tumour's nuclear-magnetism library module, for storing and providing human brain magnetic resonance tumor image and related data;
Human brain tumour's nuclear-magnetism storehouse matching module, connect described lesion segmentation module and described human brain tumour's nuclear-magnetism library module respectively, carry out mating obtaining with the related data in described human brain tumour's nuclear-magnetism library module for the related data of the tumor region by described human brain magnetic resonance tumor segmentation image and the experience segmentation threshold that image carries out blood vessel segmentation is split to described human brain magnetic resonance tumor;
Blood vessel segmentation module, connect described lesion segmentation module and described human brain tumour's nuclear-magnetism storehouse matching module respectively, for analyzing described human brain magnetic resonance tumor segmentation image and obtain blood vessel gray threshold according to described experience segmentation threshold, and described blood vessel gray threshold is adopted to carry out blood vessel segmentation to described human brain magnetic resonance tumor segmentation image;
Wherein, described intensity profile human brain magnetic resonance tumor segmentation image being analyzed to the non-tumor region specifically obtaining human brain magnetic resonance tumor segmentation image, and adopting gray probability function f (m) to represent described intensity profile, gray probability function f (m) represents that gray shade scale is the ratio of quantity shared by described non-tumor region of the pixel of m; Described experience segmentation threshold Th represent non-vascular region in described non-tumor region account for described non-tumor region ratio and according to following formulae discovery blood vessel gray threshold:
Wherein, 1≤m≤n, when calculating m=1 to m=n f (m) with threshold grey scale value when being greater than Th, and using described threshold grey scale value as described blood vessel gray threshold.
2. the magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse according to claim 1, is characterized in that, described human brain tumour's nuclear-magnetism storehouse matching module comprises further:
Tumor area matching unit, connects described lesion segmentation module and described human brain tumour's nuclear-magnetism library module respectively, for extracting the area of the tumor region of described human brain magnetic resonance tumor segmentation image and carrying out area matched with described human brain tumour's nuclear-magnetism library module;
Tumor's profiles matching unit, connect described lesion segmentation module, described human brain tumour's nuclear-magnetism library module respectively, for extracting the outline data of the tumor region of described human brain magnetic resonance tumor segmentation image and carrying out outline with described human brain tumour's nuclear-magnetism library module;
The match is successful judging unit, connect described tumor area matching unit, described tumor's profiles matching unit and described blood vessel segmentation module respectively, matched data for being obtained by described tumor area matching unit and described tumor's profiles matching unit judges whether that the match is successful, and sends to described blood vessel segmentation module by what obtain after the match is successful to the experience segmentation threshold that described human brain magnetic resonance tumor segmentation image carries out blood vessel segmentation process.
3. the magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse according to claim 1, it is characterized in that, described blood vessel segmentation module comprises further:
Blood vessel gray threshold computing unit, described blood vessel gray threshold computing unit connects described lesion segmentation module and described human brain tumour's nuclear-magnetism storehouse matching module respectively, for analyzing described human brain magnetic resonance tumor segmentation image and obtain described blood vessel gray threshold according to described experience segmentation threshold;
Vessel segmentation unit, connects described blood vessel gray threshold computing unit, carries out blood vessel segmentation for adopting described blood vessel gray threshold to described human brain magnetic resonance tumor segmentation image.
4. the magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse according to claim 1, it is characterized in that, also comprise largest connected territory computing module, described largest connected territory computing module connects described blood vessel segmentation module, for calculating the largest connected territory of the image after blood vessel segmentation.
5. the magnetic resonance image (MRI) blood vessel segmentation system based on human brain tumour's nuclear-magnetism storehouse according to claim 4, it is characterized in that, also comprise image smoothing module, largest connected territory computing module described in described image smoothing model calling, for smoothing to the image calculated behind largest connected territory.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310159783.3A CN103236062B (en) | 2013-05-03 | 2013-05-03 | Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310159783.3A CN103236062B (en) | 2013-05-03 | 2013-05-03 | Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103236062A CN103236062A (en) | 2013-08-07 |
| CN103236062B true CN103236062B (en) | 2016-01-13 |
Family
ID=48884101
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310159783.3A Expired - Fee Related CN103236062B (en) | 2013-05-03 | 2013-05-03 | Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103236062B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE112015004952T5 (en) * | 2014-10-31 | 2017-09-21 | Koninklijke Philips N.V. | Magnetic resonance imaging protocol for segmenting a picture detail |
| CN106651874B (en) * | 2016-12-08 | 2019-06-28 | 温州医科大学 | Space domain splitting method after brain tumor surgery based on multi-modal MRI data |
| CN106600599B (en) * | 2016-12-23 | 2019-12-10 | 东软集团股份有限公司 | Method and device for analyzing surrounding degree of tumor surrounding blood vessel |
| CN116152253B (en) * | 2023-04-20 | 2023-06-16 | 四川大学华西医院 | Cardiac magnetic resonance mapping quantification method, system and storage medium |
| CN116630358B (en) * | 2023-07-25 | 2023-09-26 | 潍坊医学院附属医院 | Threshold segmentation method for brain tumor CT image |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714153A (en) * | 2009-11-16 | 2010-05-26 | 杭州电子科技大学 | Visual perception based interactive mammography image searth method |
| WO2011109872A1 (en) * | 2010-03-11 | 2011-09-15 | Commonwealth Scientific And Industrial Research Organisation | Local intensity method and system for pathology detection |
| CN103069455A (en) * | 2010-07-30 | 2013-04-24 | 皇家飞利浦电子股份有限公司 | Organ-specific enhancement filter for robust segmentation of medical images |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090012382A1 (en) * | 2007-07-02 | 2009-01-08 | General Electric Company | Method and system for detection of obstructions in vasculature |
-
2013
- 2013-05-03 CN CN201310159783.3A patent/CN103236062B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101714153A (en) * | 2009-11-16 | 2010-05-26 | 杭州电子科技大学 | Visual perception based interactive mammography image searth method |
| WO2011109872A1 (en) * | 2010-03-11 | 2011-09-15 | Commonwealth Scientific And Industrial Research Organisation | Local intensity method and system for pathology detection |
| CN103069455A (en) * | 2010-07-30 | 2013-04-24 | 皇家飞利浦电子股份有限公司 | Organ-specific enhancement filter for robust segmentation of medical images |
Non-Patent Citations (1)
| Title |
|---|
| 医学图像中血管分割算法的研究;王永莉;《中国优秀硕士学位论文全文数据库信息科技辑》;20101215;第4.2.3节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103236062A (en) | 2013-08-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8965089B2 (en) | Method of analyzing a medical image | |
| Park et al. | Skull stripping based on region growing for magnetic resonance brain images | |
| CN103236062B (en) | Based on the magnetic resonance image (MRI) blood vessel segmentation system in human brain tumour's nuclear-magnetism storehouse | |
| Wong et al. | A semi-automated method for liver tumor segmentation based on 2D region growing with knowledge-based constraints | |
| CN105447872A (en) | Method for automatically identifying liver tumor type in ultrasonic image | |
| Chang et al. | A neural network for thyroid segmentation and volume estimation in CT images | |
| CN104794708A (en) | Atherosclerosis plaque composition dividing method based on multi-feature learning | |
| CN106558031B (en) | A kind of image enchancing method of the colored eyeground figure based on imaging model | |
| CN106651874B (en) | Space domain splitting method after brain tumor surgery based on multi-modal MRI data | |
| Jayade et al. | Review of brain tumor detection concept using MRI images | |
| Almi'ani et al. | Automatic segmentation algorithm for brain MRA images | |
| CN113298754B (en) | Method for detecting control points of outline of prostate tissue | |
| Mandwe et al. | Detection of brain tumor using k-means clustering | |
| Bouzidi et al. | BrainSeg3D to detect Multiple sclerosis lesions using magnetic resonance imaging | |
| CN102124471A (en) | Methods for enhancing vascular patterns in cervical imagery | |
| Kaur et al. | Detection of brain tumor using NNE approach | |
| Massich et al. | Automatic seed placement for breast lesion segmentation on US images | |
| Almi'ani et al. | A modified region growing based algorithm to vessel segmentation in magnetic resonance angiography | |
| Yao et al. | Automatic segmentation of colonic polyps in CT colonography based on knowledge-guided deformable models | |
| Pardo et al. | Automated skin lesion segmentation with kernel density estimation | |
| Katyal et al. | Gaussian intensity model with neighborhood cues for fluid-tissue categorization of multisequence MR brain images | |
| Sanchez et al. | A new system of computer-aided diagnosis of skin lesions | |
| Bhavani | Automatic MR brain tumor detection using possibilistic C-means and K-means clustering with color segmentation | |
| CN114129188B (en) | Ultrasonic contrast perfusion flow direction estimation method and imaging system | |
| Kayalvizhi et al. | Analysis of Alzheimer MR brain images using entropy based segmentation and Minkowski Functional |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C53 | Correction of patent of invention or patent application | ||
| CB02 | Change of applicant information |
Address after: 100070 No. 11 East Fengtai Road, Beijing, Fengtai District Applicant after: CRSC COMMUNICATION & INFORMATION Corp. Address before: 100070 No. 11 East Fengtai Road, Beijing, Fengtai District Applicant before: Beijing China Railway Huachen Communication Information Technology Co.,Ltd. |
|
| COR | Change of bibliographic data |
Free format text: CORRECT: APPLICANT; FROM: BEIJING CHINA RAILWAY HUACHEN COMMUNICATION INFORMATION TECHNOLOGY CO.,LTD. TO: TONGHAO COMMUNICATION INFORMATION GROUP CO., LTD. |
|
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| ASS | Succession or assignment of patent right |
Owner name: JIANG TAO Effective date: 20141212 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20141212 Address after: 100070 No. 11 East Fengtai Road, Beijing, Fengtai District Applicant after: CRSC COMMUNICATION & INFORMATION Corp. Applicant after: Jiang Tao Address before: 100070 No. 11 East Fengtai Road, Beijing, Fengtai District Applicant before: CRSC COMMUNICATION & INFORMATION Corp. |
|
| 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: 20160113 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |