CN106709925A - Method and device for locating vertebral block in medical image - Google Patents
Method and device for locating vertebral block in medical image Download PDFInfo
- Publication number
- CN106709925A CN106709925A CN201611263050.4A CN201611263050A CN106709925A CN 106709925 A CN106709925 A CN 106709925A CN 201611263050 A CN201611263050 A CN 201611263050A CN 106709925 A CN106709925 A CN 106709925A
- Authority
- CN
- China
- Prior art keywords
- candidate
- connected domain
- vertebra
- disc area
- vertebra block
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 230000004807 localization Effects 0.000 claims description 19
- 230000000877 morphologic effect Effects 0.000 claims description 13
- 238000001514 detection method Methods 0.000 claims description 12
- 230000011218 segmentation Effects 0.000 claims description 12
- 230000002708 enhancing effect Effects 0.000 claims description 8
- 210000000988 bone and bone Anatomy 0.000 claims description 7
- 230000000717 retained effect Effects 0.000 claims description 4
- 238000004519 manufacturing process Methods 0.000 claims description 3
- 108010001267 Protein Subunits Proteins 0.000 claims description 2
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 238000003745 diagnosis Methods 0.000 abstract description 3
- 238000003860 storage Methods 0.000 description 18
- 238000010586 diagram Methods 0.000 description 15
- 238000012545 processing Methods 0.000 description 11
- 210000000278 spinal cord Anatomy 0.000 description 11
- 238000003384 imaging method Methods 0.000 description 10
- 238000002595 magnetic resonance imaging Methods 0.000 description 9
- 238000004891 communication Methods 0.000 description 6
- 238000002591 computed tomography Methods 0.000 description 6
- 238000003756 stirring Methods 0.000 description 6
- 210000000577 adipose tissue Anatomy 0.000 description 4
- 238000010968 computed tomography angiography Methods 0.000 description 4
- 239000003814 drug Substances 0.000 description 4
- 210000004556 brain Anatomy 0.000 description 3
- 238000010276 construction Methods 0.000 description 3
- 238000013500 data storage Methods 0.000 description 3
- 238000002583 angiography Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 238000002224 dissection Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 210000003041 ligament Anatomy 0.000 description 2
- 238000002600 positron emission tomography Methods 0.000 description 2
- 210000001519 tissue Anatomy 0.000 description 2
- 238000003325 tomography Methods 0.000 description 2
- 208000010392 Bone Fractures Diseases 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 206010061246 Intervertebral disc degeneration Diseases 0.000 description 1
- 208000020307 Spinal disease Diseases 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 208000018180 degenerative disc disease Diseases 0.000 description 1
- 230000003412 degenerative effect Effects 0.000 description 1
- 238000000151 deposition Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 208000021600 intervertebral disc degenerative disease Diseases 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000001575 pathological effect Effects 0.000 description 1
- VYMDGNCVAMGZFE-UHFFFAOYSA-N phenylbutazonum Chemical compound O=C1C(CCCC)C(=O)N(C=2C=CC=CC=2)N1C1=CC=CC=C1 VYMDGNCVAMGZFE-UHFFFAOYSA-N 0.000 description 1
- 238000004321 preservation Methods 0.000 description 1
- 238000007639 printing Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 208000024891 symptom Diseases 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10081—Computed x-ray tomography [CT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10088—Magnetic resonance imaging [MRI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10072—Tomographic images
- G06T2207/10104—Positron emission tomography [PET]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30008—Bone
- G06T2207/30012—Spine; Backbone
Landscapes
- Magnetic Resonance Imaging Apparatus (AREA)
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention provides a method and a device for locating a vertebral block in a medical image. The method comprises the following steps of obtaining volume data comprising a plurality of layers of two-dimensional images; obtaining a candidate intervertebral disc region based on the volume data; detecting the candidate intervertebral disc region, and obtaining an intervertebral disc region; locating a vertebral block region according to the intervertebral disc region. According to the method and the device for locating the vertebral block in the medical image, provided by the invention, the vertebral block can be quickly located, the stability and the robustness are strong, and the quality of a diagnosis image is increased.
Description
【Technical field】
The present invention relates to field of medical image processing, more particularly to a kind of vertebra block of medical image localization method and its dress
Put.
【Background technology】
Backbone medical image, such as backbone MRI have very important work in various spinal diseases are diagnosed
With such as vertebra degenerative dislocation is that research intervertebral disc degeneration sexually revises most simple, accurate detection methods at present.Doctor is led to
Cross vertebra block positioning and obtain more backbone information, improve diagnostic accuracy.The full backbone splicing of such as MR images can be utilized and position vertebra
Block provides information for the gamma correction of multiple segment data；Bone analysis application can using the location information of vertebra block as vertebra block split etc.
Initialization of algorithm etc..
Mainly there be the method for vertebra block positioning in currently available technology：
A kind of method that interverbebral disc quantifies mark.The method is first using expert's point location and the method based on masterplate to ridge
Columnar region carries out coarse segmentation, then with Huffman change detection spinal cord, after according to positioning criterion, along spinal cord direction of curve,
Each interverbebral disc of whole backbone is positioned, vertebra block message is positioned based on the interverbebral disc.The method is existing by interpersonal interaction
Most clearly image is processed in image slice, is failed effectively using complete image slice information, and limited by template
There is deviation in system, the accuracy of acquired scan image.
One kind is based on curve matching, the method screened to the candidate's vertebra disk in image/vertebra block anchor point.The method vertebra
Block precise positioning and Accurate Segmentation propose that framework is more complicated, and high precision is still relatively time-consuming.
Therefore, it is necessary to propose to improve to existing vertebra block localization method.
【The content of the invention】
The technical problems to be solved by the invention are to provide the localization method and its device of vertebra block in a kind of medical image, use
In solving to position vertebra block message in all types of full spine images, to improve the quality and precision of diagnostic image.
In order to solve the above technical problems, the present invention provides a kind of localization method of vertebra block in medical image, including following step
Suddenly:
If obtaining the volume data comprising dried layer two dimensional image；
Candidate's disc area is obtained based on the volume data；
Candidate's disc area is detected, disc area is obtained；
Vertebra block region is positioned according to the disc area.
Further, it is described based on volume data data acquisition candidate's disc area, including：
If the line enhancing dried layer two dimensional image；
The enhanced two dimensional image of Threshold segmentation line, obtains first candidate's disc area on every layer of two dimensional image；
Operation is carried out out to the first candidate disc area according to first direction, the on every layer of two dimensional image is obtained
Two candidate's disc areas；
Second candidate region described in Morphological scale-space, obtains candidate's disc area of the volume data.
Further, detection candidate's disc area, obtains disc area, including：
Several connected domains are obtained according to candidate's disc area；
Based on anatomic information, tail bone region and head zone are removed in candidate's disc area；
According to removal tail bone region and the candidate spinal region area of head zone, several connected domains are obtained；
The volume of each connected domain is calculated, the connected domain is retained if the volume of connected domain is located in threshold range；
Otherwise remove the connected domain.
Further, detection candidate's disc area, obtains disc area, also includes：
Obtain first, second, and third direction of any connected domain；If connected domain third direction is flat with where first direction
The angle of plane is less than threshold value a where face and/or second direction, then retain the connected domain；
Otherwise remove the connected domain；
Wherein, first, second, and third direction is vertical two-by-two.
Further, detection candidate's disc area, obtains disc area, also includes：
The angle that the central point of any one connected domain is formed with other any two connected domain central point lines is calculated, is adopted
Connected domain where retaining central point of the angle less than threshold value b with alternative manner.
Further, the alternative manner includes：
Calculate the angle that any one connected domain central point is formed with other any two connected domain central point lines；
The frequency that any central point of the statistics angle more than threshold value b occurs, where the maximum central point of removal frequency of occurrence
Connected domain；
Repeat the above steps, until any one connected domain central point and other any two connected domain central point line shapes
Into angle be less than threshold value b.
Further, it is described that vertebra block region is positioned according to the disc area, including：
An optional connected domain in the disc area；
The set for choosing the marginal point of the connected domain constitutes the first boundary line；
With first boundary line respectively along growth length c on third direction and third direction reverse direction, obtain described
Candidate's vertebra block region of connected domain correspondence third direction and third direction reverse direction；
According to candidate's vertebra block region, the corresponding vertebra block region of the connected domain is determined.
Further, it is described according to candidate's vertebra block region, determine the corresponding vertebra block region of the connected domain, including：
The image value in candidate's vertebra block region of the connected domain correspondence third direction and third direction reverse direction is calculated respectively
Standard deviation or gradient；
Selection standard difference or the less candidate's vertebra block region of gradient are the corresponding vertebra block region of the connected domain；
The connected domain in all disc areas is traveled through, is repeated the above steps, it is each in the acquisition correspondence disc area
Several vertebra block regions of individual connected domain.
In order to solve the above technical problems, the present invention also provides a kind of positioner of vertebra block in medical image, including：
Image acquisition unit, if for obtaining the volume data comprising dried layer two dimensional image；
Candidate's interverbebral disc acquiring unit, for obtaining candidate's disc area based on the volume data；Candidate's intervertebral
Disk acquiring unit includes, enhanson, segmentation subelement, opens operation subelement and Morphological scale-space subelement；
Detection unit, for detecting candidate's disc area, obtains disc area；
Positioning unit, for positioning vertebra block region according to the disc area.
Further, the positioning unit includes：
Candidate's vertebra block obtains subelement, for inverse square respectively along second direction and second direction with first boundary line
Upward production length c, obtains candidate's vertebra block region of the connected domain correspondence second direction or second direction reverse direction；
Judgment sub-unit, the candidate's vertebra for calculating the connected domain correspondence second direction and second direction reverse direction respectively
The standard deviation of the image value in block region；The less candidate's vertebra block region of selection standard difference is the corresponding Zhui Kuai areas of the connected domain
Domain.
Present invention contrast prior art has following beneficial effect：Using human dissection information, by positioning intervertebral panel
Domain, and then realize the positioning of vertebra block；Using the partially dark interverbebral disc of inverse extra large gloomy line enhancing acquisition and the fringe region of vertebra block, it is adaptable to
Different images sequence.Make full use of backbone spatial information and form study to remove interverbebral disc false positive, improve positioning precision.On
The localization method and its device for stating the vertebra block of embodiment offer quickly position vertebra block, stability and strong robustness, improve diagnosis
The quality of image.
【Brief description of the drawings】
Fig. 1 is the positioner applied environment schematic diagram of vertebra block in the medical image in one embodiment of the invention；
Fig. 2 is the schematic diagram of the positioner of vertebra block in one embodiment of the invention traditional Chinese medicine image；
Fig. 3 is the schematic diagram of the positioner of vertebra block in further embodiment of this invention traditional Chinese medicine image；
Fig. 4 is the localization method schematic flow sheet of the vertebra block of embodiment of the present invention traditional Chinese medicine image；
Fig. 5 is T1, T2 and STIR backbone MRI schematic diagram in the embodiment of the present invention；
Fig. 6 is the method flow schematic diagram of acquisition candidate's disc area in the embodiment of the present invention；
Fig. 7 a~7d is each step structural representation of vertebra block positioning in one embodiment of the invention in the embodiment of the present invention
Result schematic diagram；
Fig. 8 is the schematic diagram of patient coordinate system in the embodiment of the present invention；
Fig. 9 is three schematic diagrames of principal direction of interverbebral disc in the embodiment of the present invention；
Figure 10 is the method flow schematic diagram of removal interverbebral disc false positive in the embodiment of the present invention；
Figure 11 is the method flow schematic diagram in positioning vertebra block region in one embodiment of the invention；
Figure 12 is the result schematic diagram in candidate's vertebra block region in one embodiment of the invention.
【Specific embodiment】
Elaborate many details in order to fully understand the present invention in the following description.But the present invention can be with
Much it is different from other manner described here to implement, those skilled in the art can be in the situation without prejudice to intension of the present invention
Under do similar popularization, therefore the present invention is not limited by following public specific implementation.
Secondly, the present invention is described in detail using schematic diagram, when the embodiment of the present invention is described in detail, for purposes of illustration only, institute
It is embodiment to state schematic diagram, and it should not limit the scope of protection of the invention herein.
In order to solve the above technical problems, the present embodiment provides the positioner place applied environment of vertebra block in medical image
Schematic diagram.The applied environment can include one or more processing systems, one or more storage single devices, one or more
Imaging system, one or more input-output apparatus, can be that distribution can also be centralization between unit and equipment
, can be it is local can also be long-range.
It is exemplary, as shown in figure 1, the system includes：Imaging system U100, storage device U200, image processing system
U300, and input-output apparatus U400.
The imaging system U100, for obtaining medical image.The medical image medical science is included but is not limited to by each
The imaging system scanning collection of class mode obtains three-dimensional or two dimensional image, if two dimensional image structure of the 3-D view comprising dried layer
Into volume data, it is also possible to by such as storing be image archiving and communication system (Picture Archiving and
Communication Systems, PACS) etc. the transmission of internal or external storage system obtain.The mode is included but is not limited to
Magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), computed tomography (CT), CT angiographic images (CTA, CT
Angiography), one or more of positron emission computerized tomography (Positron Emission Tomography, PET) etc.
Combination.Medical image can be sent to storage device U200 and make storage treatment by the imaging system U100, it is also possible to be transmitted
Image procossing is carried out to image processing system U300.
The storage device U200 can be the equipment with store function.The data that storage imaging system U100 is collected
The various data produced in (for example, medical image of imaging system shooting) and image processing system U300 work.The storage
Equipment U200 can be local, or long-range.The storage device U200 can by after information digitalization again with profit
The storage device of electricity consumption, the mode such as magnetically or optically is stored.The storage device U200 may also be used for depositing various information
Such as program and data etc..Above-mentioned storage device is to list some examples, in the present embodiment, the positioning dress of vertebra block
The storage device that can be used in the working environment put is not limited thereto.
Described image processing system U300, including the vertebra block positioner, such as Fig. 2 and Fig. 3 be used for process the doctor
Image is learned, the result in positioning vertebra block region is obtained.In the present embodiment, the positioner of the vertebra block includes image acquisition unit
U310 and candidate's interverbebral disc acquiring unit U320, detection unit U330 and positioning unit U340.
Image acquisition unit U310, if for obtaining the volume data comprising dried layer two dimensional image.The volume data can
Obtained with by the imaging system U100 or described storage devices U200.
Candidate's interverbebral disc acquiring unit U320, for for based on the volume data obtain candidate's disc area,
Including enhanson U321, segmentation subelement U322, open operation subelement U324 and Morphological scale-space subelement U324.Institute
Enhanson U321 is stated, if for the dried layer two dimensional image, successively strengthening the line-like area in two dimensional image.Such as intervertebral
Plate edge region, backbone, spinal cord etc..The segmentation subelement U322, for the enhanced two dimensional image of cut-off rule, obtains every layer
First candidate's disc area on two dimensional image；It is described to open operation subelement U323, for according to first direction to described the
One candidate's disc area carries out out operation, obtains second candidate's disc area on every layer of two dimensional image；The first party
To can be x directions (head pin direction) in patient coordinate system, described to open operation refer to retain the connected domain on x directions, corrodes y
Connected domain (such as spinal cord area) on direction, shows as only including only retaining the connected domain on x directions in two dimensional image.It is described
Morphological scale-space subelement U324, for the second candidate region described in Morphological scale-space, obtains candidate's intervertebral of the volume data
Disk area, removes the noise regions such as broken bone, spinal cord, and result is merged with the volume data, obtains the volume data
Candidate's disc area.
The positioning unit U340, for positioning vertebra block region according to the disc area, including candidate's vertebra block is obtained
Subelement U341, respectively along production length b in second direction and second direction reverse direction, obtains for first boundary line
Take candidate's vertebra block region of the connected domain correspondence second direction or second direction reverse direction；And judgment sub-unit U342, use
In the standard of the image value in the candidate's vertebra block region for calculating the connected domain correspondence second direction and second direction reverse direction respectively
Difference；The less candidate's vertebra block region of selection standard difference is the corresponding vertebra block region of the connected domain.
The input-output apparatus U400 can be to image processing system U300 input datas, it is also possible to receive at image
The data of reason system U300 outputs, the positioning result result of such as vertebra block, and the data that will be exported are with numeral, character, image, sound
The forms such as sound show.The data of output can be sent to external equipment, it is also possible to not send.The output data not sent can
To store in storage device.The input-output apparatus can include but is not limited to display device, printing device, plotter,
The combination of one or more in image output system, voice output system, magnetic recording equipment etc..In certain embodiments, have
A little external equipments can simultaneously play a part of to be input into and export, for example, desktop computer, notebook, smart mobile phone, flat board are electric
Brain, personal digital assistant (personal digital assistance, PDA) etc..
Above-mentioned image processing system U300 can be actually existed in application system, it is also possible to completed by cloud computing platform
Corresponding function.Wherein, cloud computing platform include but is not limited to storage-type cloud platform based on data storage, based on processing data
Calculation type cloud platform and take into account data storage and treatment comprehensive cloud computing platform.The cloud platform that system is used can be
Public cloud, private clound, community cloud or mixed cloud etc..For example, according to actual needs, some medical images that system is received can be with
Calculated by cloud platform and/or stored.Other medical images, can be by local diagnosis unit and/or system data
Storehouse is calculated and/or is stored.
It should be noted that imaging system U100, storage device U200, image processing system U300, input-output apparatus
Connection or communication between U400 can be wired, or wireless.
Above for the description of the positioner and its applied environment of vertebra block in medical image, only for convenience of description, not
The application can be limited within the scope of illustrated embodiment.It is appreciated that for a person skilled in the art, should understanding
After the principle of system, unit may be combined in the case of without departing substantially from this principle, or constitute subsystem
System is connected with other units, to implementing various amendments and change in the above method and systematic difference field form and details.
It is including but not limited to public cloud, privately owned for example, storage device U200 can be the cloud computing platform with data storage function
Cloud, community cloud and mixed cloud etc..Such deformation, within the protection domain of the application.
Method in order to solve quick positioning vertebra block in the prior art, the present embodiment provides a kind of vertebra block of medical image
Localization method, as shown in the localization method schematic flow sheet of the vertebra block of Fig. 4 embodiment of the present invention traditional Chinese medicine images, methods described bag
Include：
Step S410, if obtaining the volume data comprising dried layer two dimensional image；The medical image medical science is included but is not limited to
Three-dimensional or two dimensional image is obtained by the imaging system scanning collection of all kinds of mode, if two dimension of the 3-D view comprising dried layer
The volume data of image construction, it is also possible to by such as storing be image archiving and communication system (Picture Archiving and
Communication Systems, PACS) etc. the transmission of internal or external storage system obtain.The mode is included but is not limited to
Magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), computed tomography (CT), CT angiographic images (CTA, CT
Angiography), one or more of positron emission computerized tomography (Positron Emission Tomography, PET) etc.
Combination, or longitudinal relaxation (T1), transverse relaxation (T2) and quick upset to recover sequence (STIR) etc. not homotactic
MRI, the present invention is not specifically limited to this, preferred backbone MRI in the present embodiment；Described volume data is led to
Often in scan plan to obtain accurate scanning parameter and if the coarse scanning that carries out obtains that ten is several layers of or the dried layer two dimension such as tens layers
The sequence image of image construction.
Exemplary, volume data is with for the difference sequence backbone MRI such as T1, T2 and STIR in the present embodiment
Example.Adipose tissue is in highlighted state in backbone MRI, so for spine regions, the brightness of interverbebral disc under different sequences
And heterogeneity, as shown in figure 5, for T1 and T2 sequences, interverbebral disc is in dark state, in STIR sequences, interverbebral disc brightness
Differ, and for the brightness of the celiolymph near brain, for T1 sequence grays, and be in then for T2 and STIR sequences
It is pale, therefore, in the prior art, it is difficult to the positioning of vertebra block is realized in different types of medical image, it is desirable to provide Yi Zhonggai
The localization method for entering, with suitable for various types of medical images.
Step S420, candidate's disc area is obtained based on the volume data.To accelerate subsequent treatment speed, it is also possible to first
Removal background operation is done to the volume data, specifically, can be default by program, or by counting grey level histogram etc. certainly
Dynamic program provides the threshold value of self adaptation, and the image value of background parts is for example set to 0 by removal background air part.
Because human spine is formed by connecting by 33 pieces of vertebra blocks by means of ligament, joint and interverbebral disc, a yielding ridge is formed
Post cylinder.Ridge block is divided into cervical vertebra, thoracic vertebrae, lumbar vertebrae, rumpbone and coccyx.In the present embodiment, by positioning disc area, Jin Ershi
The positioning of existing vertebra block.It is as shown in Figure 6 a method flow diagram for acquisition candidate's disc area, including：
Step S421, if the line enhancing dried layer two dimensional image；As shown in figure 5, in the difference sequence ridge such as T1, T2 and STIR
In post MRI, spinal cord, adipose tissue, the brightness of interverbebral disc are simultaneously differed, and the marginal portion of interverbebral disc and vertebra block is in
Now partially dark wire result.Therefore according to the characteristic, in the present embodiment, the two dimension is successively strengthened using the enhancing of line Enhancement Method
The relatively low linear structure of image value in image, such as, based on extra large gloomy line Enhancement Method, will take Hessian matrix negative sign and successively process institute
Two dimensional image is stated, line enhancing result as shown in Figure 7a is obtained.
Step S422, the enhanced two dimensional image of Threshold segmentation line obtains first candidate's intervertebral on every layer of two dimensional image
Disk area；The enhanced two dimensional image of line that abovementioned steps are obtained, the main ridge such as including vertebra block, interverbebral disc, spinal cord of the image
The noises such as vertebra tissue, and adipose tissue, non-spinal tissue.Therefore need to obtain first candidate's intervertebral panel by Threshold segmentation
Domain, as shown in Figure 7b.The present embodiment uses threshold segmentation method, and the threshold value can be by counting the enhanced two dimensional image
In the grey level histogram of each pixel obtain, it is also possible to by the way that input-output apparatus U400 is default or image processing system
Software program in U300 determines.
Step S423, operation is carried out out to the first candidate disc area according to first direction, obtains every layer of two dimension
Second candidate's disc area on image.As shown in FIG. 8 and 9, in patient coordinate system, X-direction represents human body left and right directions,
Y-direction represents fore-and-aft direction, and Z-direction represents human body head pin direction.Because interverbebral disc is in the form of annular discs in spatial shape, according to vertebra
Plane where disk coordinate system, interverbebral disc first direction and second direction along Y-axis it is parallel or with Y-direction be sitting in reference axis angle (-
45 °, 45 °) between, therefore the present embodiment collect can along the Y direction to cross Threshold segmentation obtain first candidate's disc area
Operation is carried out out, is used to remove the wire connected domain in Z-direction, such as noise such as spinal cord, adipose tissue, on every layer of two dimensional image
Second candidate's disc area, as shown in Figure 7 c.In certain embodiments, for volume data, can be in YZ directions institute
The two dimensional image of plane, i.e. coronal-plane is by S421~S423 where the two dimensional image of plane, i.e. sagittal plane and/or XZ directions
After step process, respectively to first candidate's disc area on the two dimensional image where coronal-plane and sagittal plane along patient coordinate
X-direction in system carries out out operation, obtains the second candidate disc area as shown in Fig. 7 c and 7b.
Step S424, the second candidate region described in Morphological scale-space obtains candidate's disc area of the volume data.Through
In second candidate region crossed after above-mentioned steps treatment, also comprising series of discrete point, for example, the ridge retained in operation is opened
Marrow isolated point, it is therefore desirable to by the non-disc area in the second candidate region described in Morphological scale-space.Grasped using morphology
Make, such as the kernel function in X-direction, remove noise point, it is exemplary, because the noise point such as spinal cord is in two-dimentional sagittal view
As the length in upper X-direction is shorter, therefore given threshold, preferably threshold value are 5 millimeters, and pixel of the removal less than 5 millimeters is constituted
Region, so as to remove the obvious noise such as spinal cord, isolated point, puies forward candidate's disc area of data described in acquisition.
Perform step S430：Candidate's disc area is detected, disc area is obtained.Due to being obtained in abovementioned steps
Candidate's disc area include interverbebral disc false positive, the false positive refers to because of various reasons not possessing positive symptom
Pathological examination goes out the result of the positive, such as described interverbebral disc false positive is the non-interverbebral disc of acquisition after line enhancing and pretreatment operation
Point, such as part spinal cord point, it is therefore desirable to detect detection candidate's disc area, to remove interverbebral disc false positive,
Obtain disc area.
In the present embodiment, according to interverbebral disc in spatial relation, and architectural feature removal interverbebral disc in the form of annular discs
False positive, methods described as shown in Figure 10, comprises the following steps：
1) several connected domains are obtained according to candidate's disc area；Based on anatomic information, in candidate's intervertebral
Tail bone region and head zone are removed in disk area；In due to full backbone volume data, the top half for cervical region is general
It is head, does not have pyramidal portion, and the tail bone in the volume data of waist section part is too small, can not detect, according to these dissections
Information can remove the false positive that a part detects brain or bladder edge, the main part of preservation spinal.
2) according to removal tail bone region and the candidate spinal region area of head zone, several three-dimensional communications are obtained
Domain；The volume of each connected domain is calculated, the connected domain is retained if the volume of connected domain is located in threshold range；Otherwise remove the company
Logical domain.The connected domain for not meeting human intervertebral disc volume size is removed by this step.Exemplary, for the direction of X, Y and Z tri-
Upper resolution ratio is 1 millimeter of image, and general human intervertebral disc magnitude range should be 100 to 2000 millimeters, if the body of connected domain
Product is located in the threshold range and then retains the connected domain, otherwise removes the connected domain.
4) first, second, and third direction of any connected domain is obtained；If flat where connected domain third direction and a direction
The angle of plane is less than threshold value a where face and/or second direction, then retain the connected domain；Otherwise remove the connected domain；This step
In, by analyzing whether principal direction meets a morphological feature for vertebra disk.Because interverbebral disc is in the form of annular discs in space, and along disease
People's coordinate system Y-axis is parallel or left side is less than 45 degree or right side is downward-sloping less than 45 degree, is carried out according to the morphological feature
Judge, if connected domain third direction is less than threshold value a with the angle of plane where head pin direction (i.e. Z-direction), retain the connection
Domain, otherwise removes the connected domain.Wherein three principal directions of interverbebral disc are as shown in figure 9, first, second, and third direction two
Two vertically calculate the direction of any connected domain first, second, and third.
5) angle that any one connected domain central point is formed with other any two connected domain central point lines is calculated, is adopted
Connected domain where retaining central point of the angle less than threshold value b with alternative manner.Because interverbebral disc is presented pipe in space
Shape structure, so central point link group angle should be less than threshold value b, therefore can remove interverbebral disc by analysis center's corner dimension
False positive, by the alternative manner, removal successively is unsatisfactory for the connected domain of structural information feature, comprises the following steps：
The angle that any one connected domain central point is formed with other any two connected domain central point lines；Statistics angle
The frequency that any central point more than threshold value b occurs, the connected domain where the maximum central point of removal frequency of occurrence；Repeat above-mentioned
Step, until any one connected domain central point is less than threshold value with the angle that other any two connected domain central point lines are formed
Untill b；Wherein described threshold value b spans are less than 45 degree.
In this step S430, by the morphology synthesis of the volume to connected domain, principal direction characteristic and space tubular structure
Judge, make full use of interverbebral disc morphological characteristic to remove interverbebral disc false positive, the stability and robustness that acquisition intervertebral is made an inventory are high.
It should be noted that interverbebral disc false positive can also be removed according to interverbebral disc morphological feature by other, such as detect in space
In whether be disc structure, the present embodiment is not especially limited to this.
Step S440 is performed, vertebra block region is positioned according to the disc area.In due to human spine's construction, vertebra block is borrowed
Ligament, joint and interverbebral disc are formed by connecting, and form a yielding backbone cylinder, therefore in this step, by abovementioned steps
The disc area of acquisition, realizes the positioning of vertebra block.Exemplary, the method in the positioning vertebra block region is as shown in figure 11：
Step S441, an optional connected domain in the disc area.
Step S442~step S443, the set for choosing the marginal point of the connected domain constitutes the first boundary line；With described
First boundary line obtains the connected domain correspondence the 3rd respectively along growth length c on third direction and third direction reverse direction
Direction and candidate's vertebra block region of third direction reverse direction.In certain embodiments, first boundary line, can choose institute
One section in the set of the marginal point for stating connected domain, such as shown in Figure 12, the mesozone of the set of the marginal point can be chosen
Domain so that candidate's vertebra block region avoids growing to non-centrum region in growth course.The third direction can be as shown in Figure 9
The normal orientation of interverbebral disc disk plane, may also be the Z-direction in patient coordinate system.Patient coordinate system Z is chosen in the present embodiment
Direction, i.e. third direction.The growth length c is length of the interverbebral disc to vertebra block, is to meet other connections in growth course
Stop during domain acquired.
Step S444~step S445, according to candidate's vertebra block region, determines the corresponding vertebra block region of the connected domain.
In certain embodiments, candidate's vertebra block region of the connected domain correspondence third direction and third direction reverse direction is calculated respectively
The standard deviation of image value；The less candidate's vertebra block region of selection standard difference is the corresponding vertebra block region of the connected domain；According to vertebra
Body structure, first boundary line side Shi Zhuikuai regions, intra-zone image value is more homogeneous, and two opposite sides are interverbebral discs
With vertebra block region, the standard deviation of the image intensity value in the region is relatively large, or gradient have it is prominent more, therefore, it can basis
The size of the standard deviation of the image value in candidate's vertebra block region of third direction and third direction reverse direction, chooses less candidate's vertebra
Block region is the corresponding vertebra block region of the connected domain.The connected domain in all disc areas is traveled through, is repeated the above steps, obtained
Take several vertebra block regions of each connected domain in the correspondence disc area.By in volume data according to obtain several vertebra blocks
The highlighted vertebra block region in region, obtains the positioning result of the vertebra block such as Fig. 7.
In sum, the present invention provides the positioning and its device of vertebra block in a kind of medical image, by using in volume data
In the present embodiment, by positioning disc area, and then the positioning of vertebra block is realized；Partially dark vertebra is obtained using inverse extra large gloomy line enhancing
The fringe region of disk and vertebra block, it is adaptable to different images sequence；Backbone spatial information and form study are made full use of to remove
Interverbebral disc false positive, improves positioning precision.The localization method and its device of the vertebra block that above-described embodiment is provided quickly position vertebra block,
Stability and strong robustness, improve the quality of diagnostic image.
Although the present invention with preferred embodiment disclose as above, can not the application be limited in illustrated embodiment scope it
It is interior.It is appreciated that for a person skilled in the art, after the principle for understanding the flow, can be to the one of the flow
Or multiple operations carry out the change in details.For example, adjusting operation order, mixing operation, decoupling operation, remove one or more
Operate, increase one or more operations etc..These change all without departing from scope of the claims.
Claims (10)
1. in a kind of medical image vertebra block localization method, it is characterised in that comprise the following steps:
If obtaining the volume data comprising dried layer two dimensional image；
Candidate's disc area is obtained based on the volume data；
Candidate's disc area is detected, disc area is obtained；
Vertebra block region is positioned according to the disc area.
2. the localization method of vertebra block as claimed in claim 1, it is characterised in that described that candidate's vertebra is obtained based on the volume data
Between disk area, including：
If the line enhancing dried layer two dimensional image；
The enhanced two dimensional image of Threshold segmentation line, obtains first candidate's disc area on every layer of two dimensional image；
Operation is carried out out to the first candidate disc area according to first direction, the second time on every layer of two dimensional image is obtained
Select disc area；
Second candidate region described in Morphological scale-space, obtains candidate's disc area of the volume data.
3. the localization method of vertebra block as claimed in claim 1, it is characterised in that detection candidate's disc area,
Disc area is obtained, including：
Several connected domains are obtained according to candidate's disc area；
Based on anatomic information, tail bone region and head zone are removed in candidate's disc area, obtain candidate's regiones vertebralis
Domain area；；
According to candidate spinal region area, several connected domains are obtained；
The volume of each connected domain is calculated, the connected domain is retained if the volume of connected domain is located in threshold range；
Otherwise remove the connected domain.
4. the localization method of vertebra block as claimed in claim 3, detection candidate's disc area,
Disc area is obtained, is also included：
Obtain first, second, and third direction of any connected domain；If third direction and plane and/or second where first direction
The angle of plane is less than threshold value a where direction, then retain the connected domain；
Otherwise remove the connected domain；
Wherein, first, second, and third direction is vertical two-by-two.
5. the localization method of vertebra block as claimed in claim 3, detection candidate's disc area, obtain intervertebral panel
Domain, also includes：
The angle that the central point of any one connected domain is formed with other any two connected domain central point lines is calculated, using repeatedly
Connected domain where retaining central point of the angle less than threshold value b for method.
6. the localization method of vertebra block as claimed in claim 5, it is characterised in that the alternative manner includes：
Calculate the angle that any one connected domain central point is formed with other any two connected domain central point lines；
The frequency that any central point of the statistics angle more than threshold value b occurs, the company where the maximum central point of removal frequency of occurrence
Logical domain；
Repeat the above steps, until any one connected domain central point and the formation of other any two connected domain central point lines
Angle is less than threshold value b.
7. the localization method of vertebra block as claimed in claim 1, it is characterised in that described that vertebra is positioned according to the disc area
Block region, including：
An optional connected domain in the disc area；
The set for choosing the marginal point of the connected domain constitutes the first boundary line；
With first boundary line respectively along growth length c on third direction and third direction reverse direction,
Obtain candidate's vertebra block region of the connected domain correspondence third direction and third direction reverse direction；
According to candidate's vertebra block region, the corresponding vertebra block region of the connected domain is determined.
8. the localization method of vertebra block as claimed in claim 7, it is characterised in that described according to candidate's vertebra block region, really
The execution following step S1-S3 in the corresponding vertebra block region of the connected domain, including iteration calmly, obtains the correspondence disc area
In each connected domain several vertebra block regions：
S1, calculates the image value in candidate's vertebra block region of the connected domain correspondence third direction and third direction reverse direction respectively
Standard deviation or gradient；
S2, selection standard difference or the less candidate's vertebra block region of gradient are the corresponding vertebra block region of the connected domain；
S3, travels through the connected domain in all disc areas.
9. the positioner of the vertebra block of a kind of medical image, it is characterised in that including：
Image acquisition unit, if for obtaining the volume data comprising dried layer two dimensional image；
Candidate's interverbebral disc acquiring unit, for obtaining candidate's disc area based on the volume data；Candidate's intervertebral is examined and seized
Taking unit includes, enhanson, segmentation subelement, open operation subelement and Morphological scale-space subelement；
Detection unit, for detecting candidate's disc area, obtains disc area；
Positioning unit, for positioning vertebra block region according to the disc area.
10. the positioner of vertebra block as claimed in claim 8, it is characterised in that the positioning unit includes：
Candidate's vertebra block obtain subelement, for first boundary line respectively along second direction and second direction reverse direction
Production length c, obtains candidate's vertebra block region of the connected domain correspondence second direction or second direction reverse direction；
Judgment sub-unit, the candidate Zhui Kuai areas for calculating the connected domain correspondence second direction and second direction reverse direction respectively
The standard deviation of the image value in domain；The less candidate's vertebra block region of selection standard difference is the corresponding vertebra block region of the connected domain.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611263050.4A CN106709925A (en) | 2016-12-30 | 2016-12-30 | Method and device for locating vertebral block in medical image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611263050.4A CN106709925A (en) | 2016-12-30 | 2016-12-30 | Method and device for locating vertebral block in medical image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106709925A true CN106709925A (en) | 2017-05-24 |
Family
ID=58906388
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611263050.4A Pending CN106709925A (en) | 2016-12-30 | 2016-12-30 | Method and device for locating vertebral block in medical image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106709925A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108053400A (en) * | 2017-12-21 | 2018-05-18 | 上海联影医疗科技有限公司 | Image processing method and device |
| CN112132837A (en) * | 2020-08-19 | 2020-12-25 | 心医国际数字医疗系统(大连)有限公司 | A method, system, electronic device and storage medium for automatic extraction of thoracic bones |
| CN113240661A (en) * | 2021-05-31 | 2021-08-10 | 平安科技（深圳）有限公司 | Deep learning-based lumbar vertebra analysis method, device, equipment and storage medium |
| CN114187320A (en) * | 2021-12-14 | 2022-03-15 | 北京柏惠维康科技有限公司 | Spine CT image segmentation method and spine imaging identification method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060110017A1 (en) * | 2004-11-25 | 2006-05-25 | Chung Yuan Christian University | Method for spinal disease diagnosis based on image analysis of unaligned transversal slices |
| US20080077372A1 (en) * | 2001-08-17 | 2008-03-27 | Bioniche Life Sciences Inc. | Conformation-activity relationship of apoptosis-inducing phosphodiester oligonucleotides |
| CN102663820B (en) * | 2012-04-28 | 2014-10-22 | 清华大学 | Three-dimensional head model reconstruction method |
| CN104851107A (en) * | 2015-06-08 | 2015-08-19 | 武汉联影医疗科技有限公司 | Vertebra positioning method based on CT sequence image |
| CN105405129A (en) * | 2015-10-31 | 2016-03-16 | 上海联影医疗科技有限公司 | Reconstruction method and device of medical image |
-
2016
- 2016-12-30 CN CN201611263050.4A patent/CN106709925A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080077372A1 (en) * | 2001-08-17 | 2008-03-27 | Bioniche Life Sciences Inc. | Conformation-activity relationship of apoptosis-inducing phosphodiester oligonucleotides |
| US20060110017A1 (en) * | 2004-11-25 | 2006-05-25 | Chung Yuan Christian University | Method for spinal disease diagnosis based on image analysis of unaligned transversal slices |
| CN102663820B (en) * | 2012-04-28 | 2014-10-22 | 清华大学 | Three-dimensional head model reconstruction method |
| CN104851107A (en) * | 2015-06-08 | 2015-08-19 | 武汉联影医疗科技有限公司 | Vertebra positioning method based on CT sequence image |
| CN105405129A (en) * | 2015-10-31 | 2016-03-16 | 上海联影医疗科技有限公司 | Reconstruction method and device of medical image |
Non-Patent Citations (1)
| Title |
|---|
| 杨帆: "《数字图像处理与分析》", 31 May 2015, 北京航空航天大学出版社 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108053400A (en) * | 2017-12-21 | 2018-05-18 | 上海联影医疗科技有限公司 | Image processing method and device |
| CN108053400B (en) * | 2017-12-21 | 2021-06-15 | 上海联影医疗科技股份有限公司 | Image processing method and device |
| US11129582B2 (en) | 2017-12-21 | 2021-09-28 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for medical imaging of intervertebral discs |
| US11826191B2 (en) | 2017-12-21 | 2023-11-28 | Shanghai United Imaging Healthcare Co., Ltd. | System and method for medical imaging of intervertebral discs |
| CN112132837A (en) * | 2020-08-19 | 2020-12-25 | 心医国际数字医疗系统(大连)有限公司 | A method, system, electronic device and storage medium for automatic extraction of thoracic bones |
| CN112132837B (en) * | 2020-08-19 | 2024-07-26 | 心医国际数字医疗系统(大连)有限公司 | Chest bone automatic extraction method, system, electronic equipment and storage medium |
| CN113240661A (en) * | 2021-05-31 | 2021-08-10 | 平安科技（深圳）有限公司 | Deep learning-based lumbar vertebra analysis method, device, equipment and storage medium |
| CN113240661B (en) * | 2021-05-31 | 2023-09-26 | 平安科技（深圳）有限公司 | Deep learning-based lumbar vertebra bone analysis method, device, equipment and storage medium |
| CN114187320A (en) * | 2021-12-14 | 2022-03-15 | 北京柏惠维康科技有限公司 | Spine CT image segmentation method and spine imaging identification method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106683090B (en) | The localization method and its system of rib cage in medical image | |
| Li et al. | Lung nodule detection with deep learning in 3D thoracic MR images | |
| CN106600609B (en) | Spine segmentation method and system in medical image | |
| Hawkes | Algorithms for radiological image registration and their clinical application | |
| US8588486B2 (en) | Apparatus and method for isolating a region in an image | |
| CN111340767B (en) | Brain tumor scalp positioning image processing method and system | |
| CN109034221A (en) | A kind of processing method and its device of cervical cytology characteristics of image | |
| US12039717B2 (en) | Region dividing device, method, and program, similarity determining apparatus, method, and program, and feature quantity deriving apparatus, method, and program | |
| CN111429474A (en) | Mammary gland DCE-MRI image focus segmentation model establishment and segmentation method based on mixed convolution | |
| CN114974575A (en) | Breast cancer neoadjuvant chemotherapy curative effect prediction device based on multi-feature fusion | |
| JP2006314778A (en) | Medical image processing apparatus and medical image processing method | |
| JP2013545520A (en) | Image search engine | |
| CN106709925A (en) | Method and device for locating vertebral block in medical image | |
| CN106709920B (en) | Blood vessel extraction method and device | |
| JP3234668U (en) | Image recognition system for scoliosis by X-ray | |
| US20110052024A1 (en) | method of analysing stroke images | |
| CN105405129A (en) | Reconstruction method and device of medical image | |
| Hess et al. | Deep learning for multi-tissue segmentation and fully automatic personalized biomechanical models from BACPAC clinical lumbar spine MRI | |
| CA2778599C (en) | Bone imagery segmentation method and apparatus | |
| CN116958059A (en) | A computer-aided reference system and method for merging multi-modal breast images | |
| Dourthe et al. | Automated segmentation of spinal muscles from upright open MRI using a multiscale pyramid 2D convolutional neural network | |
| CN116433976A (en) | Image processing method, device, equipment and storage medium | |
| CN104915989B (en) | Blood vessel three-dimensional dividing method based on CT images | |
| CN109063208A (en) | A kind of medical image search method merging various features information | |
| Salehi et al. | Automatic diagnosis of disc herniation in two-dimensional MR images with combination of distinct features using machine learning methods |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170524 |