CN105326524A - Medical imaging method and device capable of reducing artifacts in image - Google Patents
Medical imaging method and device capable of reducing artifacts in image Download PDFInfo
- Publication number
- CN105326524A CN105326524A CN201410371170.0A CN201410371170A CN105326524A CN 105326524 A CN105326524 A CN 105326524A CN 201410371170 A CN201410371170 A CN 201410371170A CN 105326524 A CN105326524 A CN 105326524A
- Authority
- CN
- China
- Prior art keywords
- image
- filtering
- high attenuation
- produce
- pixel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
- Image Processing (AREA)
Abstract
The invention discloses a medical imaging device and a medical imaging device capable of reducing artifacts in an image, and relates to a medical imaging method. The method comprises the following steps: scanning an object including a background part and a high attenuation part in one or more different scanning directions, so that plural of projected images are generated; and for each of the projected images, filtering the projected image so as to generate an original filtered image; shrinking pixel points in the high attenuation part, close to boundary, of the projected image inwards the high attenuation part in the scanning direction or in a direction opposite to the scanning direction, and filling a pixel gray value, close to the adjacent background part, in the original region of the high attenuation part which shrinks inwards, so that a shrunk image is generated; filtering the shrunk image so as to a shrunk filtered image; fusing the original filtered image and the shrunk filtered image, so that a filtered image with undershoot reduced is generated; and projecting the filtered image with undershoot reduced in an opposite direction so as to obtain a reconstructed image. In addition, the invention also relates to a medical imaging device for implementing the method.
Description
Technical field
The present invention relates to a kind of method and apparatus of medical imaging, particularly, relate to a kind of medical imaging procedure and device of the artifact reduced in image.
Background technology
Medical Imaging Technology can obtain the image of the internal structure of patient or object to be seen when not needing physical penetration skin, be thus widely used clinically.Projection Radiographic technique (projectionradiography) is a kind of conventional Medical Imaging Technology, and its characteristic based on X-ray (that is, having different absorbances in different materials), utilizes X-ray to produce two dimensional image.Such as, chromatography imaging technique (tomosynthesis) is a kind of like this projection Radiographic technique, it adopts different physical principles, if X-ray is through the differential transmission (differentialtransmission) during target object, obtain view data, build tomographic map (such as, the interior three-dimensional of human body or other structures to be imaged is reproduced).
Some limitations of chromatography imaging technique or restriction may cause producing artifact or other defect in reconstruction image, such as, the high attenuation portions with high X-ray absorbance in patient body or on health, as metallic material parts (such as prosthese, implant, screw, pin etc.) may cause forming artifact in tomographic process.For example, using back projection (backprojection) as in the tomographic process of algorithm for reconstructing, there is the object of high X-ray absorbance as metal, the artifact producing the copy as this metal on different rebuilding planes can be caused, that is, ripple (ripple) artifact.Artifact will seriously reduce the quality of final image, sometimes even causes image cannot be used for diagnosis, and therefore the artifact reduced in image has very important effect.
Summary of the invention
Embodiments of the invention relate in one aspect to a kind of medical imaging procedure.In the method, scan to produce a plurality of projected image along one or more different scanning direction to object, wherein said object comprises background parts and high attenuation portions.Following operation is carried out to each projected image: filtering is carried out to projected image, to produce original filtration image; By the pixel on the border of the close high attenuation portions in the high attenuation portions in projected image and background parts along the opposite direction of scanning direction or scanning direction towards high attenuation portions internal contraction, and the grey scale pixel value in the former high attenuation portions region at the described original place of high attenuation portions of internally shrinking is replaced to the grey scale pixel value being adjacent background parts grey scale pixel value and being close, to produce contractible graph picture; Filtering is carried out to described contractible graph picture, to produce contraction filtering image; Merge described original filtration image again and shrink filtering image to produce the filtering image of undershoot minimizing.The filtering image that a plurality of undershoots that a plurality of projected image described in back projection produces reduce, to obtain reconstruction image.
The another aspect of embodiments of the invention relates to a kind of being used for the medical imaging apparatus of object imaging.This device comprises radiographic source and ray detector, data collecting system, processor and display device.Described ray detector passes by the ray of the object scanned after being used for receiving and being sent by described radiographic source, and described radiographic source and ray detector are used for scanning along one or more different scanning direction the object comprising background parts and high attenuation portions.Described data collecting system is used for receiving from the data of described ray detector when scanning object, to produce corresponding data for projection.Described processor is used for receiving the data for projection from described data collecting system, to produce projected image, in order to rebuild the image of described object.This processor comprises contraction module, filtration module, Fusion Module and back projection module, wherein, shrink module to be used for the pixel on the border of the close high attenuation portions in the high attenuation portions in described projected image and background parts along the opposite direction of scanning direction or scanning direction towards high attenuation portions internal contraction, and the grey scale pixel value in the former high attenuation portions region at the described original place of high attenuation portions of internally shrinking is replaced to the grey scale pixel value being adjacent background parts grey scale pixel value and being close, to produce contractible graph picture; Filtration module, is used for carrying out filtering to produce original filtration image to described projected image, carries out filtering shrink filtering image to produce to described contractible graph picture; Fusion Module, is used for merging described original filtration image and shrinking filtering image to produce the filtering image of undershoot minimizing; Back projection module, is used for the filtering image that undershoot described in back projection reduces, to obtain reconstruction image.Described display device shows the image of described acquisition by responding described processor.
Accompanying drawing explanation
Be described for embodiments of the invention in conjunction with the drawings, the present invention may be better understood, in the accompanying drawings:
Fig. 1 is the three-dimensional block diagram of the digital radiographic chromatographic imaging system in an exemplary embodiment, and this system disposition has a table top.
Fig. 2 is the three-dimensional block diagram of the digital radiographic chromatographic imaging system in an exemplary embodiment, and this system disposition has a metope.
Fig. 3 A shows a kind of filter kernel in the spatial domain in an embodiment, and Fig. 3 B shows this spatial domain grey scale pixel value curve, Fig. 3 C show the filter kernel shown in Fig. 3 A be applied to carry out filtering in the curve shown in Fig. 3 B after the curve that obtains.
Fig. 4 shows the flow chart of the formation method of one embodiment of the invention.
Fig. 5 A and 5B respectively illustrate original projection image corresponding in an embodiment one group and high attenuation portions inwardly shrink after projected image at the area filling shunk after the grey scale pixel value of approximate background part.
Solid line thinner in Fig. 6 and dotted line to show respectively in an embodiment one group of corresponding original filtration image I
foriginalwith contraction filtering image I
fshrunkengrey scale pixel value curve near the border of high attenuation portions and background parts, thicker solid line illustrates I by certain way
foriginaland I
fshrunkenthe filtering image I eliminating undershoot artifact obtained after merging
uSfreegrey scale pixel value curve near this border.
Fig. 7 A and 7B to respectively illustrate in an embodiment one group of corresponding original filtration image and eliminates the filtering image of undershoot artifact.
Fig. 8 show schematically show the situation of in an embodiment, metallic object being carried out to back projection.
Fig. 9 shows the flow chart of the formation method of another embodiment of the present invention.
Figure 10 A and 10B to respectively illustrate in an embodiment one group of corresponding original filtration image and eliminates the filtering image of ripple artifact.
Detailed description of the invention
Below will be described specific embodiments of the invention.Unless otherwise defined, the technical term used in claims and description or scientific terminology should be in the technical field of the invention the ordinary meaning that the personage with general technical ability understands." first ", " second " that use in patent application specification of the present invention and claims and similar word do not represent any order, quantity or importance, and are only used to distinguish different ingredients.The similar word such as " one " or " one " does not represent that quantity limits, but represents to there is at least one."or" refers to any or all that comprise in cited element or object." comprise " or the similar word such as " comprising " mean to appear at " comprising " or " comprising " before element or object contain the element or object and equivalent element thereof that appear at " comprising " or " comprising " presented hereinafter, do not get rid of other elements or object." connection " or " being connected " etc. similar word be not defined in physics or the connection of machinery, no matter but can comprise electric connection, be direct or indirectly.The language of approximation used herein can be used for quantitative expression, shows to allow quantity to have certain variation when not changing basic function.Therefore, this exact value itself is not limited to the numerical value that the language such as " approximately ", " left and right " is revised.In some cases, approximating language may be relevant with the precision of gauge.
Provide a kind of formation method in the embodiment of the present invention, it obtains the medical image of target object by chromatographic imaging system.In the method, back-projection algorithm can be adopted, filter back-projection algorithm scheduling algorithm rebuilds the image of tomography, in filtering by interior move and the method for filling border to reduce or eliminate artifact in the plane such as undershoot (undershooting) and upper punch (overshooting), by applying the method for different weight for different pixels point to reduce or eliminate the outer artifact of the planes such as ripple in back projection process, by these hands sections, the existence due to the contour decayed object of metal of rebuilding in image and the artifact defect produced can be reduced or eliminated.
Methods described herein are widely used in various tomographic reconstruction system.Because these methods are specially adapted to various X-ray chromatography imaging technique, as radiochromatograph imaging technique or digital object chromatography imaging technique (radordigitalobjecttomosynthesis), for convenience of description, herein by mainly in conjunction with a kind of as applicant in JIUYUE in 2007 application on the 10th the U.S. Patent application US2009/0003679 being entitled as " Systemandmethodforadigitalx-rayradiographictomosynthesis userinterface " described in chromatographic imaging system exemplary discussion and description are carried out to the method for the embodiment of the present invention, but should understand, method of the present invention is also applicable to other chromatographic imaging systems or other tomographic reconstruction systems.
Fig. 1 and 2 respectively illustrates the example of the embodiment of digital radiographic chromatographic imaging system 100,200.System 100,200 comprises digital chromatographic imaging system 120,220.Fig. 1 shows a kind of version with platform, and it comprises the X-ray detector 104 under the x-ray source 102 being contained in mechanism 160 and the table top 118 being positioned at platform 116.Fig. 2 shows a kind of version with metope, and it comprises the x-ray source 202 being contained in mechanism 260 and the X-ray detector 204 be arranged on metope 216.Under the patient of examine is placed in the radiation of the X-ray beam 108,208 emitted by x-ray source 102,202, described X-ray beam 108,208 is mapped on the body of patient 106,206 of examine, some of be radiated through or walk around patient arrive and collision detector 104,204.
In one embodiment, described x-ray source 102,202 can be X-ray tube, and the patient 106,206 of described examine can be the non-life body of the mankind, animal, test body mould (testphantom) and (or) other examine.
The patient 106 or 206 of described examine is placed between described x-ray source 102,202 and described detector 104,204.In chromatography process, as illustrated in fig. 1 and 2, described x-ray source 102,202 moves along plane 110,210, also rotates, to guarantee X-ray beam 108,208 alignment detector 104,204 all the time in this process simultaneously.As previously mentioned, described x-ray source 102,202 normally along with the plane 112 of described detector 104,204,212 parallel plane 110,210 movements, although it also may shift out outside plane, it is parallel to described detector 104,204 substantially.When obtaining actinogram, described detector 104,204 can remain on a fixing position.In the process of x-ray source 102,202 along plane 110,210 movement, discrete diverse location can obtain by described detector 104,204 projection radiography (projection view) of the series of discrete of patient 106,206.Projecting image data can be obtained, for rebuilding sectioning image from described projection view.
In described digital radiographic tomographic process, x-ray source 102,202 is rotated by arc or linear movement and aim in the scanning once in certain angular range 114,214 (scanning angle) that fixing detector 104,204 completes, and includes a series of low dosage exposure (lowdoseexposures).The scanning once carried out from different projection angles, described x-ray source 102,202 sends multiple exposure.Described scanning angle 114,214 refers to the angle from the first projection exposure to last projection exposure.Described scanning angle 114,214 is usually in the scope of 20 to 60 degree.
In one embodiment, described detector 104,204 can comprise a series of detecting element, usual respective pixel, be used for perception through or walk around the intensity of X-ray of patient, and produce the electronic signal of the intensity of the incident X-rays bundle represented on each detecting element.These electronic signals obtained are processed, is used for the three-dimensional volumetric images of anatomical structure of reconstruction patients.The intensity of the X-ray of each detecting element of described shock can change, and depends on the decay of X-ray and intervenes the absorbance of structure.
Fig. 1 and 2 show schematically show the computer workstation 130,230 that is connected to described digital chromatographic imaging system 120,220 further, which provide user interface 140,240, be used for setting at least one reconstruction, dosage for digital radiographic tomographic process as herein described and (or) gathering the parameter of acquisition aspect.
Described digital chromatographic imaging system 120,220 can be used for obtaining and process projecting image data, and the volumetric image of reconstruction patients or 3-D view.Described digital chromatographic imaging system 120,220 arranges to obtain projecting image data, and processes these data to carry out and check and analyze.
Described computer workstation 130,230 comprises at least one computer 132,232, this computer have controller 134,234, processor 136,236, memorizer 138,238 and user interface 140,240.Described processor 136,236 can be connected to described controller 134,234, described memorizer 138,238 and described user interface 140,240.User 150,250 controls the operation of described digital radiographic chromatographic imaging system 100,200 by computer workstation 130,230.In one embodiment, described memorizer 138,238 can be storage device, memory plane, data storage and any other file layout known in the art.
Described digital chromatographic imaging system 120,220 is controlled by controller 134,234, this controller not only can provide energy, also can provide the control signal of digital chromatography checking sequence (digitaltomosynthesisexaminationsequences), this signal comprises the position of x-ray source relative to patient and detector.Described controller 134,234 obtains the signal produced in detector by instruction, also can perform the function of various signal processing and filtering, such as, to the initial adjustment of dynamic range, the staggered of Digital Image Data etc.Usually, described controller 134,234, by digital chromatographic imaging system 120,220 described in command operating, performs inspection scheme, processes the data obtained.In one embodiment, described controller 134,234 receives the instruction from computer 132,232.In one embodiment, described controller 134,234 can be a part for described digital chromatographic imaging system 120,220, replaces described computer workstation 130,230.In one embodiment, described computer 132,232 comprises or is connected to described user interface 140,240, come to carry out reciprocal action with user 150,250, to select and (or) to change and clinical relevant parameter, (rebuilding setting) and acquisition parameter etc. are arranged in such as dosage, section.In one embodiment, described digital radiographic chromatographic imaging system 100,200 is by downloading or be integrated in the making for realizing running of software in computer 132,232 or algorithm.
In one embodiment, described user interface 140,240 is visual interfaces, it can comprise a plurality of predefined instrument, these instruments allow user 150,250 to check, select and edit reconstruction parameter (setting), check and selective dose parameter, and check, select and edit chromatography parameter.Described user interface 140,240 also allows described user 150,250 to check the image of reconstruction.
In one embodiment, described user interface 140,240 can comprise at least one input equipment, the information on the display being used for inputting and (or) select user interface 140,240 on described a plurality of predefined instruments of showing.In one embodiment, at least one input equipment described can be touch screen, mouse, keyboard, at least one button or other any input equipment known in the art.
In use, patient 106,206 is in certain position, obtain by detector 104,204 to make the projection view of target object (part for such as health), after obtaining projecting image data from described projection view, can obtain a plurality of sectioning images of described target object based on described projecting image data, the 3-dimensional image as this object represents.In practice, described a plurality of sectioning image obtains by different algorithm for reconstructing, includes but not limited to back-projection algorithm and filter back-projection algorithm.
Undertaken by described system in the process of imaging, can scan object along different scanning angles and direction, to obtain the projected image of different angles, then by diverse ways, described projected image is being rebuild, the reconstruction image required for acquisition.In certain embodiments, described object comprises high attenuation portions and background parts.In this article, high attenuation portions refers to the part with high decay characteristics in object to be scanned or on it.Described high attenuation portions may than the remainder of object to the absorbance of X-ray, and namely background parts is much higher, and such as, high attenuation portions is to the absorbance of X-ray in the distance of about 1 centimetre, and at least 2 times or 3 times are to background parts; Or after absorption of x-rays, the grey scale pixel value of high attenuation portions is at least little than the grey scale pixel value of its neighborhood by 20 or about 10.Typically, described high attenuation portions relates to high density material, such as metal, tooth filling, skeleton etc.Especially, described high attenuation portions to relate in clinical practice in patients with implantation body or the parts installed on a patient body, such as, be used for correcting the parts of skeleton, it may comprise the object relating to metal of various clinical practice widely, as being used for correcting the prostheses such as the screw of skeleton and nail etc. and tooth, spinal column, femur and knee.And background parts typically refers to the tissue of patient.Especially, high attenuation portions is metal and background parts is nonmetal.
When rebuilding the described projected image comprising the object of high attenuation portions and background parts with filter back-projection algorithm, undershoot or upper punch may be there is on the border of high attenuation portions and background parts.
Filter back-projection algorithm comprises filtering and back projection.Typically, filtering is a kind of derivation operation, can help the contrast and the acutance that strengthen image.Normally, filtering is a kind of process that will tiltedly become filter kernel and be multiplied with window function:
Wherein, filter kernel kernel (t) wherein the heart (t=0 place) be on the occasion of, other positions are negative value.At very strong border (border as between the metal parts implanted and bodily tissue) place, widely different between x (i) and x (i-1) or x (i+1), in this case, although kernel (1) or kernel (-1) is more much smaller than kernel (0), Xc (i) still can be very large by the impact of x (i-1) or x (i+1), thus there will be undershoot or upper punch.
Such as, Fig. 3 A shows the example of a kind of oblique change filter kernel in spatial domain, and Fig. 3 B shows this spatial domain grey scale pixel value curve, and higher platform phase wherein represents the contour attenuation portions of metal, and all the other represent background parts.The convolution process of described filtering normally in spatial domain, in this process, kernel is as shown in Figure 3A applied in curve as shown in Figure 3 B.Curve is as shown in Figure 3 C obtained after filtering.But, due to high attenuation portions as metal has high X-ray absorbance, its adjacent edges may be caused to produce fierce Strength Changes, and this produces the problem of undershoot or upper punch after may causing filtering, and upper punch or undershoot will cause occurring artifact rebuilding in image.Such as; as shown in Figure 7 A; in reconstruction image; there is the artifact defect (as illustrated in the drawing black surround 701) because undershoot or upper punch cause in usual meeting some comparison of light and shade boundaries near high attenuation portions (as implanted metal parts) and human tissue structure junction, the existence of artifact hampers the observation to these junctions.Because projected image normally to carry out filtering, so undershoot artifact generally produces along this scanning direction along the scanning direction of x-ray source.
In embodiment in one aspect of the invention, reduce the artifacts such as undershoot by following methods.As shown in Figure 4, in step 41, carry out after scanning creates a plurality of projected image along one or more different scanning direction to the object comprising background parts and high attenuation portions, when described object being rebuild based on described projected image filter back-projection algorithm, in step 42, following operation is carried out: in step 421, to projected image (I for each projected image
original) carry out filtering, to produce original filtration image (I
foriginal); In step 422, by in the high attenuation portions in projected image near the pixel on border along the opposite direction of scanning direction or scanning direction to high attenuation portions internal contraction, and the region at high attenuation portions place originally in described inside contraction, namely, the grey scale pixel value of grey scale pixel value close to (equal or close) of the background parts be adjacent is filled, to produce contractible graph picture (I in former high attenuation portions region
shrunken); In step 423, filtering is carried out to described contractible graph picture, shrink filtering image (I to produce
fshrunken); In step 424, merge described original filtration image and shrink filtering image to produce the filtering image (I of undershoot minimizing
uSfree).Like this, the filtering image of a corresponding undershoot minimizing all can be obtained for each projected image.At step 43, the filtering image reduced based on these undershoots carries out back projection, just can obtain the reconstruction image reduce or eliminating undershoot.Relative to original projected image, in contractible graph picture, because the high attenuation portions near border inwardly shrinks and is filled with the grey scale pixel value close to background parts wherein, the border of high attenuation portions and background parts is mobile to high attenuation portions inside, thus based on the undershoot defect of carrying out contractible graph picture in the contraction filtering image of filtering generation also to high attenuation portions internal blas, and no longer there is undershoot at original boundary, therefore, if in a certain way described original filtration image and contracting filtering image are merged, replace with the corresponding region in described contraction filtering image to make the border near zone creating undershoot artifact in original filtration image, just the filtering image not only eliminating undershoot artifact but also farthest remain original structure feature can be obtained.
Based on the characteristic of filtering, in some specific embodiments, determine the width of undershoot artifact region by gradient intensity, thus determine the distance (contracted length) that in high attenuation portions, each pixel should inwardly shrink.Suppose that scanning direction is y-axis direction, image gradient in a scanning direction can represent with dy:
Wherein, f
yfor the radioscopic image of input.So, pixel (i, j) contracted length L (i, j) calculates by following equation:
L(i，j)＝H(dy(i，j)，kernel)，
Wherein, H is the function of contracted length L (i, j) to gradient dy (i, j), is the function of the filter kernel kernel in gradient dy (i, j) and process of reconstruction.The contracted length of the pixel that gradient is larger is larger.For the pixel that gradient is very little, its contracted length, close to 0, does not shrink substantially.Comparatively large in the gradient of the pixel of strong borderline region, its contracted length is also larger.
In some specific embodiments, can estimate with the grey scale pixel value of the background parts be adjacent at the grey scale pixel value of the high attenuation portions middle filling in region (former high attenuation portions region) at place originally of described inside contraction.Especially, in a specific embodiment, the grey scale pixel value of filling in described former high attenuation portions region is the meansigma methods of the grey scale pixel value of the background parts be adjacent.
Fig. 5 A and 5B respectively illustrates original projection image (I corresponding in an embodiment one group
original) and contractible graph picture (I
shrunken).Can find out, compared with original projection image, in contractible graph picture, high attenuation portions (in figure black part) there occurs contraction along scanning direction A, that is, in contractible graph picture, high attenuation portions size is in a scanning direction less than the high attenuation portions size in a scanning direction in original projection image.
In some specific embodiments, by following equation by original filtration image I
foriginalwith contraction filtering image I
fshrunkenmerge the filtering image I reduced to produce undershoot
uSfree:
I
USfree＝(1-mask)*I
fshrunken+mask*I
foriginal，
Wherein, mask (" mask ") refers to a kind of dot chart, and certain area is wherein 1 and all the other regions are 0, and it can be used to be multiplied with target image obtain filtering image.In a specific embodiment, in mask on corresponding scanning direction from (i, j) to (i+/-C*L, the value in the region of pixel j) is 0, the value in all the other regions is 1, and wherein L is the contracted length of pixel (i, j), C is constant parameter, its correspond to described in be used for carrying out projected image and contractible graph picture the kernel of filtering.At pixel (i, j) place, in a zonule in a scanning direction, I can be found out
foriginaland I
fshrunkenbetween the minimum part of difference, and the starting point of mask mask used as pixel (i, j), along scanning oppositely from this point in high attenuation portions until within the scope of the contracted length of pixel (i, j) mask be 0.
What solid line 601 thinner in Fig. 6 and dotted line 602 represented respectively is in an embodiment one group of corresponding original filtration image I
foriginalwith contraction filtering image I
fshrunkengrey scale pixel value curve near the border of high attenuation portions and background parts, what described thicker solid line 603 represented is by I by aforesaid equation
foriginaland I
fshrunkenthe filtering image I obtained after merging
uSfreegrey scale pixel value curve near described border, wherein higher platform phase and lower platform phase represent high attenuation portions and background parts respectively, and except part near the border of high attenuation portions and background parts, 601,602 roughly overlap with 603.Can find out, the part of downward projection is had between platform phase higher in 601 and 602 and lower platform phase, the part of this downward projection represents the undershoot artifact of boundary, then there is no undershoot artifact in the grey scale pixel value curve 603 obtained after merging, thus definition and the acutance on the border of high attenuation portions can be kept.
Fig. 7 A and 7B respectively illustrates original filtering image (I in an embodiment
foriginal) and utilize described method to cut down the filtering image (I of undershoot artifact
uSfree).In the image shown in Fig. 7 A, the boundary of high attenuation portions has a black surround 701, i.e. undershoot artifact, and there is not such artifact in image shown in figure 7b, thus more clearly can show the organizational structure near border.
On the other hand, due to the existence of high attenuation portions, also ripple artifact may may be produced in the process of back projection.Fig. 8 show schematically show the process of the projected image of the object containing high attenuation portions (below for metal) and background parts being carried out to back projection.When carrying out back projection, the drop shadow curve (as L1 and L2) of traceable various angle, obtains the respective pixel on detector, at reconstructed positions superposition.Have three kinds of regions at a three dimensions, their reconstruction section is different by the influence degree of the metal in object.Wherein the first is the metallic region at metal place, and all projection views through this region are loaded with metal information; The second is impregnable background area, that is, in any projection view all not by the non-metallic regions of metal barrier, all projection views through this region do not comprise metal information, and the pixel in this region does not have artifact; The third is the background area affected by metal, namely can by the non-metallic regions of metal barrier in some projection view.If rebuild this region by the projection view of the projection view of metal barrier based on comprising these, can the information of metal be brought into this part background area, thus produce ripple artifact.Such as, as shown in Figure 10 A, there is the ripple as metal false shadow overlapping with tissue in image.The appearance of described ripple artifact reduces the contrast of object construction, and what reduce the subtle anomalies structure at metal/structural interface place can discrimination degree, may cause wrong diagnosis.
In embodiment in another aspect of this invention, reduce ripple artifact by following methods.As shown in Figure 9, in step 91, along one or more different scanning direction, scanning is carried out to the object comprising background parts and high attenuation portions and create a plurality of projected image.In step 92, following operation is carried out to each projected image: in step 921, filtering is carried out to projected image, to produce filtering image; In step 922, calculate the weight of each pixel in projected image, to produce weight map picture; In step 923, described filtering image is multiplied by mutually with weight map picture and produces weighted filtering image.In step 93, back projection is carried out to the weighted filtering image that described a plurality of projected image produces, temporarily rebuilds image to produce.In step 94, back projection is carried out to the weight map picture of described a plurality of projected image, rebuild image to produce weight.In step 95, the gray value of each pixel temporarily rebuild in image being rebuild the gray value of pixel corresponding in image divided by weight, just can obtaining required reconstruction image for exporting.
Figure 10 A and 10B to respectively illustrate in an embodiment original filtering image and utilizes described method to cut down the filtering image of ripple artifact.In the image shown in Figure 10 A, around high attenuation portions, there are some ripple artifacts 901, and there is not such artifact in image shown in fig. 1 ob, thus more clearly can show the organizational structure near border.
In certain embodiments, can reduce undershoot artifact and ripple artifact by the combination of above-mentioned two kinds of methods, that is, in the filtering step, be projected image I simultaneously
foriginalin each pixel calculate weight, to produce weight map picture, and the filtering image (I described undershoot reduced
uSfree) be multiplied by with described weight map picture the weighted filtering image producing undershoot and reduce mutually; In the step of back projection, the weighted filtering image that the undershoot that described in back projection, multiple projected image produces reduces, temporarily rebuilds image to produce; The weight map picture of multiple projected image described in back projection, rebuilds image to produce weight; The gray value of each pixel temporarily rebuild in image being rebuild the gray value of pixel corresponding in image divided by weight, obtaining the reconstruction image for exporting.The reconstruction image obtained by this method does not have undershoot artifact and ripple artifact, can clear and show exactly by scanning object organizational structure.
In one embodiment, described weight map picture is
wherein e is natural Exponents, and p is the gray value of pixel (i, j), t and s is the parameter of the one or more exposure parameters in corresponding scanning step.
Method described in previous embodiment can be cut down artifact that the existence due to high attenuation portions causes and do not reduced picture quality, other artifacts or defect is not introduced yet, therefore described method can in the original artifact region more CONSTRUCTED SPECIFICATIONs of reduction and feature, help to improve distinguishable degree, thus improve the accuracy of diagnosis.
The present invention can summarize with other the concrete form without prejudice to spirit of the present invention or principal character.Therefore, no matter from which point, above-mentioned embodiment of the present invention all can only be thought explanation of the present invention and can not limit the present invention, scope of the present invention is defined by claims, instead of to be defined by above-mentioned, therefore, any change in the implication suitable with claims of the present invention and scope, all should think to be included in the scope of claims.
Claims (10)
1. a medical imaging procedure, it comprises:
Scan to produce a plurality of projected image along one or more different scanning direction to object, wherein said object comprises background parts and high attenuation portions;
For projected image described in each:
Filtering is carried out to projected image, to produce original filtration image;
By the pixel on the border of the close high attenuation portions in the high attenuation portions in projected image and background parts along the opposite direction of scanning direction or scanning direction towards high attenuation portions internal contraction, and the grey scale pixel value in the former high attenuation portions region at the described original place of high attenuation portions of internally shrinking is replaced to the grey scale pixel value being adjacent background parts and being close, to produce contractible graph picture;
Filtering is carried out to described contractible graph picture, to produce contraction filtering image; And
Merge described original filtration image and shrink filtering image to produce the filtering image of undershoot minimizing; And
The filtering image that the undershoot that described in back projection, a plurality of projected image produces reduces, to obtain reconstruction image.
2. the method for claim 1, wherein
For projected image described in each, it comprises further:
For each pixel in described projected image calculates weight, to produce weight map picture; And the filtering image described undershoot to be reduced is multiplied by with described weight map picture the weighted filtering image producing undershoot and reduce mutually;
The filtering image that the undershoot that described in back projection, a plurality of projected image produces reduces comprises:
The weighted filtering image that the undershoot that described in back projection, a plurality of projected image produces reduces, temporarily rebuilds image to produce;
The weight map picture of a plurality of projected image described in back projection, rebuilds image to produce weight; And
The gray value of each pixel temporarily rebuild in image being rebuild the gray value of pixel corresponding in image divided by weight, obtaining the reconstruction image for exporting.
3. be used for a medical imaging apparatus for object imaging, it comprises:
Radiographic source and ray detector, wherein said ray detector passes by the ray of the object scanned after being used for receiving and being sent by described radiographic source, and described radiographic source and ray detector are used for scanning along one or more different scanning direction the object comprising background parts and high attenuation portions;
Data collecting system, is used for receiving from the data of described ray detector when scanning object, to produce corresponding data for projection;
Processor, be used for receiving the data for projection from described data collecting system, to produce projected image, in order to rebuild the image of described object, it comprises:
Shrink module, be used for the pixel on the border of the close high attenuation portions in the high attenuation portions in described projected image and background parts along the opposite direction of scanning direction or scanning direction towards high attenuation portions internal contraction, and the grey scale pixel value in the former high attenuation portions region at the described original place of high attenuation portions of internally shrinking is replaced to the grey scale pixel value being adjacent background parts and being close, to produce contractible graph picture;
Filtration module, is used for carrying out filtering to produce original filtration image to described projected image, carries out filtering shrink filtering image to produce to described contractible graph picture;
Fusion Module, is used for merging described original filtration image and shrinking filtering image to produce the filtering image of undershoot minimizing; And
Back projection module, is used for the filtering image that undershoot described in back projection reduces, to obtain reconstruction image; And
Display device, shows the image of described acquisition by responding described processor.
4. device as claimed in claim 3, wherein, described processor comprises further:
Weighting block, be used for as each pixel in described projected image calculates weight, to produce weight map picture, and the filtering image described undershoot reduced is multiplied by mutually with described weight map picture and produces weighted filtering image, and
Described back projection module is used for:
The weighted filtering image that the undershoot that described in back projection, a plurality of projected image produces reduces, temporarily rebuilds image to produce;
The weight map picture of a plurality of projected image described in back projection, rebuilds image to produce weight; And
The gray value of each pixel temporarily rebuild in image being rebuild the gray value of pixel corresponding in image divided by weight, obtaining the reconstruction image for exporting.
5. method as claimed in claim 2 or device as claimed in claim 4, wherein, described weight map picture is
wherein e is natural Exponents, and p (i, j) is the gray value of pixel (i, j), t and s is the parameter of the one or more exposure parameters in corresponding scanning step.
6. the method for claim 1 or device as claimed in claim 3, wherein, described high attenuation portions comprises metal, and described background parts comprises nonmetal.
7. the method for claim 1 or device as claimed in claim 3, wherein, described projected image and contractible graph similarly are carry out filtering by rebuilding kernel.
8. the method for claim 1 or device as claimed in claim 3, wherein, the contracted length of each pixel of the inside contraction in described high attenuation portions is determined by its gradient intensity.
9. method as claimed in claim 8 or device, wherein, pixel (i, j) contracted length L (i, j) be by equation L (i, j)=H (i, j), kernel) determine, wherein dy (i, j) be pixel (i, j) gray value gradient in a scanning direction, H is one and is used for mapping gradient dy (i, j) to contracted length L (i, j) function, kernel is used to the reconstruction kernel described projected image and contractible graph picture being carried out to filtering.
10. the method for claim 1 or device as claimed in claim 3, wherein, the filtering image I that described undershoot reduces
uSfreedescribed original filtration image I
foriginalwith contraction filtering image I
fshrunkenby equation I
uSfree=(1-mask) * I
fshrunken+ madk*I
foriginalmerge and produce, wherein, mask is a kind of dot chart, in this dot chart on corresponding scanning direction from (i, j) to (i+/-C*L, the value in the region of pixel j) is 0, the contracted length of the value in all the other regions to be 1, L be pixel (i, j), C is constant parameter, and it is corresponding to the reconstruction kernel being used for carrying out described projected image and contractible graph picture filtering.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410371170.0A CN105326524B (en) | 2014-07-31 | 2014-07-31 | The medical imaging procedure and device of the artifact in image can be reduced |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410371170.0A CN105326524B (en) | 2014-07-31 | 2014-07-31 | The medical imaging procedure and device of the artifact in image can be reduced |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105326524A true CN105326524A (en) | 2016-02-17 |
| CN105326524B CN105326524B (en) | 2018-10-26 |
Family
ID=55277449
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410371170.0A Active CN105326524B (en) | 2014-07-31 | 2014-07-31 | The medical imaging procedure and device of the artifact in image can be reduced |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105326524B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107492132A (en) * | 2017-08-31 | 2017-12-19 | 南方医科大学 | A kind of image artifacts minimizing technology of digital chromatographic imaging system |
| CN107714069A (en) * | 2016-08-10 | 2018-02-23 | 南京普爱医疗设备股份有限公司 | A kind of method of testing of the metal artifacts based on CBCT images |
| CN107784677A (en) * | 2017-10-20 | 2018-03-09 | 上海联影医疗科技有限公司 | Radioscopic image processing method, device, system and storage medium |
| CN108024794A (en) * | 2015-09-29 | 2018-05-11 | 富士胶片株式会社 | The signal processing method of optoacoustic measuring arrangement and optoacoustic measuring arrangement |
| CN111462266A (en) * | 2020-03-20 | 2020-07-28 | 北京东软医疗设备有限公司 | Image reconstruction method and device, CT (computed tomography) equipment and CT system |
| CN113470137A (en) * | 2021-06-30 | 2021-10-01 | 天津大学 | IVOCT image guide wire artifact removing method based on gray-scale weighting |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101809622A (en) * | 2007-06-15 | 2010-08-18 | 皇家飞利浦电子股份有限公司 | Apparatus for determining a high density region in an image |
| CN102521853A (en) * | 2011-12-08 | 2012-06-27 | 苏州生物医学工程技术研究所 | Reconstruction algorithm for back projection weight cone-beam CT (Computed Tomography) |
| CN103582456A (en) * | 2011-05-31 | 2014-02-12 | 株式会社岛津制作所 | Radiation tomographic image generation method and radiation tomographic image generation program |
-
2014
- 2014-07-31 CN CN201410371170.0A patent/CN105326524B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101809622A (en) * | 2007-06-15 | 2010-08-18 | 皇家飞利浦电子股份有限公司 | Apparatus for determining a high density region in an image |
| CN103582456A (en) * | 2011-05-31 | 2014-02-12 | 株式会社岛津制作所 | Radiation tomographic image generation method and radiation tomographic image generation program |
| CN102521853A (en) * | 2011-12-08 | 2012-06-27 | 苏州生物医学工程技术研究所 | Reconstruction algorithm for back projection weight cone-beam CT (Computed Tomography) |
Non-Patent Citations (1)
| Title |
|---|
| ZHAOXIA ZHANG等: "METAL ARTIFACTS REDUCTION FOR TOMOSYNTHESIS", 《2014 IEEE 11TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108024794A (en) * | 2015-09-29 | 2018-05-11 | 富士胶片株式会社 | The signal processing method of optoacoustic measuring arrangement and optoacoustic measuring arrangement |
| US11083376B2 (en) | 2015-09-29 | 2021-08-10 | Fujifilm Corporation | Photoacoustic measurement device and signal processing method of photoacoustic measurement device |
| CN107714069A (en) * | 2016-08-10 | 2018-02-23 | 南京普爱医疗设备股份有限公司 | A kind of method of testing of the metal artifacts based on CBCT images |
| CN107492132A (en) * | 2017-08-31 | 2017-12-19 | 南方医科大学 | A kind of image artifacts minimizing technology of digital chromatographic imaging system |
| CN107784677A (en) * | 2017-10-20 | 2018-03-09 | 上海联影医疗科技有限公司 | Radioscopic image processing method, device, system and storage medium |
| CN111462266A (en) * | 2020-03-20 | 2020-07-28 | 北京东软医疗设备有限公司 | Image reconstruction method and device, CT (computed tomography) equipment and CT system |
| CN111462266B (en) * | 2020-03-20 | 2023-08-04 | 北京东软医疗设备有限公司 | Image reconstruction method, device, CT equipment and CT system |
| CN113470137A (en) * | 2021-06-30 | 2021-10-01 | 天津大学 | IVOCT image guide wire artifact removing method based on gray-scale weighting |
| CN113470137B (en) * | 2021-06-30 | 2022-04-29 | 天津大学 | IVOCT image guide wire artifact removing method based on gray-scale weighting |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105326524B (en) | 2018-10-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105326524A (en) | Medical imaging method and device capable of reducing artifacts in image | |
| CN110337673B (en) | Method and system for estimating data in tomographic reconstruction based on deep learning | |
| US10939887B2 (en) | System and method of artifact correction in 3D imaging | |
| JP4409043B2 (en) | Computer program and apparatus for reconstructing image data acquired by tomosynthesis X-ray imaging system | |
| US7761136B2 (en) | Medical image processing apparatus for scanning based on a set three-dimensional region of interest | |
| CN102697514B (en) | Optimal viewing angle is selected to dissect structures visualization and patient skin dose to optimize | |
| EP1522045B1 (en) | Motion artifact correction of tomographical images | |
| US20210056688A1 (en) | Using deep learning to reduce metal artifacts | |
| US20180070902A1 (en) | Apparatus and method for 4d x-ray imaging | |
| US9437018B2 (en) | Method and apparatus for reducing artifacts in computed tomography image reconstruction | |
| CN104323789B (en) | A kind of CT scan image rebuilding method and CT scanner | |
| US8045774B2 (en) | Method for creating material-selective volume images | |
| KR101105494B1 (en) | A reconstruction method of patient-customized 3-D human bone model | |
| US10034648B2 (en) | System and method for motion artifacts reduction | |
| DE102013220663A1 (en) | Reconstruction of image data using contour data | |
| WO2019208037A1 (en) | Image analysis method, segmentation method, bone density measurement method, learning model creation method, and image creation device | |
| JP2007097977A (en) | X-ray ct apparatus | |
| CN111260748A (en) | Digital synthesis X-ray tomography method based on neural network | |
| CN109313816B (en) | Depth enhanced tomosynthesis reconstruction | |
| JP7120662B2 (en) | System and method for low-dose multispectral X-ray tomography | |
| CN115209808A (en) | Learning model creation method, image generation method, and image processing device | |
| CN113226181B (en) | Method for calibrating X-ray projection geometry in X-ray cone beam computed tomography | |
| CN113164154B (en) | Method and device for producing a panoramic tomographic image of an object to be recorded | |
| CN104545962B (en) | The medical imaging procedure and system of artifact in image can be reduced | |
| JP4982091B2 (en) | Radiation tomography equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |