CN108324324A - It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system - Google Patents
It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system Download PDFInfo
- Publication number
- CN108324324A CN108324324A CN201810201405.XA CN201810201405A CN108324324A CN 108324324 A CN108324324 A CN 108324324A CN 201810201405 A CN201810201405 A CN 201810201405A CN 108324324 A CN108324324 A CN 108324324A
- Authority
- CN
- China
- Prior art keywords
- dimensional
- track
- imaging
- pixel
- contrast
- 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
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/08—Detecting organic movements or changes, e.g. tumours, cysts, swellings
- A61B8/0891—Detecting organic movements or changes, e.g. tumours, cysts, swellings for diagnosis of blood vessels
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/13—Tomography
- A61B8/14—Echo-tomography
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
- A61B8/4483—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device characterised by features of the ultrasound transducer
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/46—Ultrasonic, sonic or infrasonic diagnostic devices with special arrangements for interfacing with the operator or the patient
- A61B8/461—Displaying means of special interest
- A61B8/466—Displaying means of special interest adapted to display 3D data
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/5269—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving detection or reduction of artifacts
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Radiology & Medical Imaging (AREA)
- Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Veterinary Medicine (AREA)
- Medical Informatics (AREA)
- Biophysics (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Vascular Medicine (AREA)
- Computer Graphics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Gynecology & Obstetrics (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
A kind of ultrasonic low frequency of present invention offer is through cranial capacity super-resolution three-dimensional contrast imaging method and system, volume imaging technique is applied to through cranium blood vessel imaging field, emit plane of ultrasound wave to encephalic using two dimensional surface energy converter, it is scanned in orthogonal both direction, scan image realizes that the zonule three-dimensional super-resolution rate through cranium blood vessel is imaged using voxel arest neighbors method and based on Markov chain Monte-Carlo multi-target tracking algorithm, obtains the Real-time High Resolution rate three-dimensional imaging to skull internal structure.
Description
Technical field
The invention belongs to ultrasound detection and ultrasonic imaging technique fields, and in particular to ultrasonic two dimensional surface transducer array,
And through cranium low frequency super-resolution volume three-D imaging method.
Background technology
Application ultrasound is mainly transcranial Doppler technology to the monitoring means in skull at present, i.e., by echo-signal
The anti-relative motion pushed away between supersonic source and scattering or reflectance target of frequency shift information, for detecting the blood flow of encephalic brain bottom major arteries
Dynamics and blood flow physiological parameter.However transcranial Doppler technology is only able to display out the two-dimensional signal of vascular flow.
Compared with tradition is through cranium two-dimensional imaging technique, there are many advantages through cranium 3 Dimension Image Technique：First, three-dimensional imaging
It is more intuitive, the three-dimensional information of Intracranial structure can be directly displayed out；Secondly, tomography chromatographic analysis is carried out to 3-D view, it can be from
The angle that conventional imaging method cannot achieve is observed, and the clinical anatomical structure and disease feelings for fully understanding detection position is facilitated
Condition；Again, 3 Dimension Image Technique can also provide more medical informations, such as tissue surface product, tissue volume, be examined for clinic
Disconnected treatment provides more accurately and reliably information.It is prior, cerebrovascular distribution imaging and relevant disease diagnosis it is desirable that
Therefore the three-dimensional information of encephalic is the inexorable trend of transcranial imaging development through cranium 3 Dimension Image Technique.
3-D supersonic imaging is also referred to as ultrasound volume imaging, according to the difference of ultrasonic three-dimensional volume data acquisition modes, Ke Yifen
It is two kinds：One is quiescent imagings, i.e., obtain the 2D signal of aerial cross sectional by Mechanical course scanning probe, then two dimension is believed
Number carry out space reconstruction obtain 3-D view, the method has higher resolution ratio, but is lost temporal information to a certain extent,
And need accurately positioning system；Second is dynamic imaging, i.e. volume imaging technique, and probe emission space is accumulated using face battle array
Ultrasonic beam directly obtains three-D ultrasonic volume data, not only has two-dimentional color ultrasound energy converter repertoire, also have three-dimensional imaging,
Image cutting, image rotation and high flat image analytic function.It is simple and quick that the advantages of volume imaging technique, is that image obtains,
But disadvantage is the limitation due to existing ultrasonic device port number, in order to be compatible with existing ultrasonic device, therefore face battle array probe size
It is smaller, cause limited view, image resolution ratio also lower than two-dimensional ultrasonic image resolution ratio.Current volume imaging is mainly used in
Three-dimensional imaging to organs such as heart, uterus, bladders does not have detection and observation to intracranial tissue also.
In conclusion how while ensureing real-time, the resolution ratio for effectively improving TCD,transcranial Doppler imaging is also one
Very crucial difficult point.In the world, French Langevin laboratory is based on random optical and rebuilds microscope imaging technology, develops connection
The ultrasound of contrast microbubbles is closed through cranium super-resolution imaging method, there are three features：The first, object of experiment is relatively thin big of skull
Mouse, thus use high frequency ultrasound imaging rather than low frequency；The second, location tracking is carried out using low concentration, single microvesicle；Third makes
It is synthesized with frame data up to ten thousand, it is larger to obtain the time.These factors make this technology face problems in Clinical practice.And
Research group of biomedical engineering system of the Institute of Technology of Israel is developed available on based on super-resolution optical fluctuation imaging method
In high concentration, the low joint microcapsular ultrasound super-resolution imaging method for obtaining the time, but its resolution ratio is tracked compared to single microvesicle
Method has apparent deficiency, and it is not suitable for the velocity imaging of progress microvesicle/blood flow.Prior, these technologies are both for two
Dimension imaging, and for ultrasound through cranial capacity three-dimensional imaging, there is not specific super-resolution imaging technology also.
Therefore, how on the basis of existing supersonic imaging apparatus, and ultra-radio frequency data acquisition device is combined, is made for combining
The shadow TCD,transcranial Doppler cerebrovascular is distributed, and dynamic 3 D ultrasonic imaging technique is applied to cranium brain, while retaining real-time, is proposed
A kind of zonule improves the resolution ratio being imaged through cranial capacity through cranium capacity of blood vessel three-dimensional imaging algorithm, is urgently to be resolved hurrily ask
Topic.
Invention content
It is an object of the invention to propose a kind of ultrasonic low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and be
Volume imaging technique is applied to, through cranium blood vessel imaging field, realize the real time three-dimensional imaging to skull internal structure, together by system
When, it is realized through cranium blood vessel using voxel arest neighbors method (VNN) and three-dimensional Markov chain Monte-Carlo multi-target tracking algorithm
The imaging of zonule three-dimensional super-resolution rate.
To achieve the goals above, present invention employs following technical schemes：
It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method, include the following steps：
1) two dimensional surface energy converter is utilized to emit plane of ultrasound wave to encephalic, the plane of ultrasound wave is in orthogonal two
It is scanned on a direction, by receiving the echo information of the scanning, obtains continuous N frames ultrasonic contrast echo rf data；
2) two dimensional beam synthesis is carried out to the echo rf data, obtains the rf data B after Beam synthesisn(n=
1,...,N)；
3) according to the rf data after Beam synthesis, the Spatial Rules volume pixel network W of each frame imaging is determinedA×B×C,
Wherein, A is x-axis direction number of pixels, and B is z-axis direction number of pixels, and C is y-axis direction number of pixels；
4) to each volume pixel point wa,b,c(a=1 ..., A；B=1 ..., B；C=1 ..., C), abbreviation tissue points,
All pixels point on all scan images is traversed in the imaging of each frame, calculates each pixel of scan image apart from the Spatial Rules
Volume pixel network WA×B×CDistance, with apart from volume pixel point wa,b,cNearest pixel value is to volume pixel point wa,b,cInto
Row interpolation；It is obtained after interpolation and Spatial Rules volume pixel network WA×B×CThe identical three-D volumes pixel network W ' of volumeA×B×C；
5) the three-D volumes pixel network W ' that will be obtainedA×B×CIt is divided into N along y-axis1Width two dimensional image Bm(m=1 ...,
N1), N1For the two dimensional surface energy converter secondary axes array number, to Bm(m=1 ..., N1) in per piece image N frame images Bm,n
(n=1 ..., N) wherein continuous N is chosen successivelymedFrame obtains BM, %n(%n=n-Nmed+1..., n), to BM, %nWith in time domain
Value filtering function is handled, and background image G is obtainedm,n：
Wherein, Distance () indicates that distance between pixels, i indicate that pixel abscissa, j indicate pixel ordinate；
6) from original image Bm,nIn subtract background image Gm,nObtain foreground image Fm,n；
7) to foreground image Fm,nAfter carrying out artefact removal and local maximum processing, by N1Width foreground image is according to substance
Product pixel network WA×B×CVolume and distribution splicing synthesize three-dimensional data, obtain n-th frame image, each frame that then will be obtained
Middle μnPosition of a pixel as contrast microbubbles, records the position coordinates of contrast microbubbles
And corresponding velocity information
8) by all μ in n-th frame imagenThe combinatorial coordinates of a pixel, obtainBy all foregrounds
The Y that image procossing obtainsnCombination, obtains
9) set distance range (≤1mm) will be less than distance range centered on each contrast microbubbles position in Y
There may be the link contrast microbubbles positions of track for contrast microbubbles position centered on other contrast microbubbles positions are default；
10) contrast microbubbles track is deduced using Markov chain Monte-Carlo multi-target tracking；
11) the final track collection of deduction is imaged.
Preferably, the two dimensional surface energy converter use 128 or 256 array elements, main shaft array element center spacing be 1mm~
1.4mm, main shaft effective aperture are 16mm~22.4mm, and secondary axes array element center spacing is 1mm~1.4mm, and secondary axes effective aperture is
16mm~22.4mm, centre frequency are 1.8MHz~2.2MHz, and the pulse width of the energy converter is less than or equal to 10 μ s, with roomy
In equal to 60%, adjacent array element crosstalk is less than or equal to -30dB, and backing decaying is less than or equal to -60dB.
Preferably, in the step 1), imaging frame rate Frt is 0.8kHz~1.2kHz, to meet the transient state of contrast microbubbles
Physical message variation requires.
Preferably, the two dimensional beam synthesis uses the minimum variance adaptive beam composition algorithm in feature based space
(for example, three-dimensional broad beam zonule Quick air imaging method of ultrasonic two-dimensional array).
Preferably, in the step 4), the nearest pixel value should be apart from volume pixel point wa,b,cNearest scanning
Tissue points (the tissue points, that is, volume in plane (plane of scanning motion is the width two dimensional image plane in each frame 3-D scanning image)
Pixel) projection pixel.The threshold quantity h of distance can be added during realizationmaxIf the tissue points with it is corresponding most
The distance of the close plane of scanning motion is more than or equal to preset threshold quantity, then illustrates that the tissue points farther out, are not required to from free scan image
Row interpolation (volume pixel point value is 0) is clicked through to the volume pixel, can so reduce the time complexity of algorithm.
Preferably, the artefact removal includes the following steps：
7.1) dimensional Gaussian kernel function is built, by two-dimensional convolution to foreground image Fm,nFiltering, obtains filtered foreground
Image F 'm,n：
Wherein, naIndicate transverse direction (x-axis) size of dimensional Gaussian kernel function, nbIndicate the longitudinal direction (z of dimensional Gaussian kernel function
Axis) size, dimensional Gaussian kernel function is：
Wherein, Δ z and Δ x is respectively the Pixel Dimensions of image axially and transversely,σxAnd σzIt is two
The variance for tieing up gaussian kernel function, is determined by the size of contrast microbubbles in the picture, can rule of thumb and using simulation imaging be determined
The size；
7.2) given threshold fthr, remove F'm,nIn be less than fthrPixel value, after being handled by local maximum, with μnIt is a
Position of the pixel as contrast microbubbles records its coordinate, and corresponding speed letter can be obtained according to position spacing and time difference
Breath；
Preferably, the threshold value fthrIt, can be by filtering further removal tail for -20dB~-15dB.
Preferably, the step 10) specifically includes following steps：
10.1) according to the deduction state of extraction, change the contrast microbubbles position in Y, and deduce obtain track collection ω=
[ω1 ω2 … ωM], wherein M is total number of tracks；
10.2) time domain iterative state equation is constructed：
Wherein A is process matrix：
Δ t=1/Frt, Frt are imaging frame rate,For the process variance matrix of Gaussian distributed；
WhereinIndicate that track concentrates radiography micro-
The coordinate information of bubble,Indicate that the velocity information of contrast microbubbles is concentrated in track, k=1,2 ... K, K are track
It is imaging frame number to concentrate contrast microbubbles number, n=1,2 ... N, N, and footmark x and z indicate horizontal and vertical coordinate respectively；
10.3) according to Kalman filtering, the microbubble locations of prediction are expressed as：
Wherein, For the observational variance matrix of Gaussian distributed,Uniformly to divide
Cloth；
10.4) statistics calculates final on trajectory number zt, current track sum nt, newborn track starting point number at, it is selected and is
The target points d of trackt, false-alarm targets sum ft, and calculate from the new trace number c derived from last iterationtAnd it is not selected
For the target points g of trackt：
ct=nt-zt-at
gt=nt-dt
10.5) Prior Probability under current deduction state is calculated：
Wherein, pzAnd pdIndicate that target termination or disappearance probability and target are selected the probability for tracing point, λ respectivelybAnd λf
The probability value and false-alarm probability value of target point new life in ideal trajectory are indicated respectively；
10.6) assume false-alarm number and deduce path to obey and be uniformly distributed, calculate based on the current track collection and respectively of deducing
The likelihood value of trajectory target points：
Wherein, M is to deduce total number of tracks, τkCurrently to deduce target point sum in track,For target
The likelihood value of point,WithObservation predicted vector and observational variance prediction matrix in Kalman filtering are indicated respectively；
10.7) posterior probability values under current deduction state are calculated：
Wherein, P (Y) indicates the prior probability of detection contrast microbubbles position；
10.8) it currently to deduce track collection ω as certain markovian state, is calculated according to Maximun Posterior Probability Estimation Method
Receive current deduction probability：
Wherein, ω ' is previous deduction track collection, and ω is this deduction track collection (currently deducing track collection)；
10.9) it generates and obeys equally distributed random number ξ, 0<ξ<1, if ξ<A (ω ', ω) then receives this deduction, will
New track collection ω replaces old track collection ω '；Otherwise, refuse this deduction, track collection ω ' is remained unchanged；
10.10) step 10.1) -10.9 is repeated), until completing iterations.
Preferably, the value of the N is 80~200, and it is 1500~3000 to deduce iterations L, to reach steady as possible
State process.
Preferably, the deduction state includes new life, extinction, update, fracture, fusion, continuation, reduction and conversion；It deduces
The extraction of state includes the following steps：It calculates track sum M under iterative process and chooses newborn state if M=0；If M=1,
Then one kind is randomly selected in other deduction states in addition to transition status；If M>1, then randomly select one in each deduction state
Kind.
A kind of ultrasound low frequency is through cranial capacity super-resolution three-dimensional angiographic imaging system, including low frequency is through the special plane of ultrasound of cranium
Energy converter and echo signal reception module, Cerebral vessels ultrasound contrast imaging module, contrast microbubbles extraction module, artefact remove mould
Module and super-resolution imaging module are deduced in block, contrast microbubbles track；Wherein：
The ultrasound echo signal that the echo signal reception module is used to receive the plane of ultrasound energy converter is adopted
Sample is simultaneously converted into rf data；
The Cerebral vessels ultrasound contrast imaging module is used for according to rf data, using two dimensional beam synthesis and the body of pixel
Plain arest neighbors method generates Cerebral vessels ultrasound contrastographic picture, and is converted into the form of Spatial Rules volume pixel network；
The contrast microbubbles extraction module is generated for removing in Cerebral vessels ultrasound contrastographic picture by time domain medium filtering
Background information, and retain the image information of contrast microbubbles；
The artefact removal module is used to filter out the motion track of the non-contrast microbubbles in the image for retaining contrast microbubbles；
Module is deduced for micro- to radiography by Markov chain Monte-Carlo multi-target tracking in the contrast microbubbles track
It is deduced bubble track；
The super-resolution imaging module be used for according to deduce obtained track collection carry out cerebral angiography super-resolution at
Picture, can be greatly improved the resolution ratio of cerebral angiography imaging, and provide reliable blood flow information.
Preferably, the low frequency uses above-mentioned two dimensional surface energy converter through the special plane of ultrasound energy converter of cranium.
Beneficial effects of the present invention are embodied in：
VNN algorithm for reconstructing is combined by the present invention with based on Markov chain Monte-Carlo multi-target tracking, can be effective
Raising TCD,transcranial Doppler cerebral angiography three-dimensional imaging resolution ratio, can reach the super-resolution of micron order or pixel class resolution ratio at
Picture, therefore can effectively reflect the distribution situation of encephalic capillary.
Further, two dimensional surface energy converter of the present invention can be mutual by carrying out in the small area of temporal bone
The scanning of vertical both direction, and then whole echoes in single pass can be collected when being observed in a certain range
Information realizes real-time three-dimensional imaging.It has the characteristics that bilateral scanning, real-time, low frequency, high-resolution and cell domain imaging.
Further, the 3-D view recovery algorithms that the present invention uses is the voxel arest neighbors method (VNN) of pixel and three
Wiki, can be effective in the TCD,transcranial Doppler cerebral angiography super-resolution imaging method of Markov chain Monte-Carlo multi-target tracking
Improve the resolution ratio being imaged through cranial capacity.
Description of the drawings
Fig. 1 is that the array element for the low frequency TCD,transcranial Doppler face array transducer that embodiment is enumerated arranges schematic diagram.
Fig. 2 is volume transducer scans schematic diagram.
Fig. 3 is skull active position schematic diagram when TCD,transcranial Doppler is imaged.
Fig. 4 is TCD,transcranial Doppler volume imaging experiment system and flow chart.
Fig. 5 is volume energy converter three-dimensional imaging flow chart.
Fig. 6 is three-dimensional deduction status diagram.
Fig. 7 is two dimension based on Markov chain Monte-Carlo multi-target tracking algorithm obtained using track collection it is final super
Resolution imaging result；(a) it is two-dimensional super-resolution rate imaging results；(b) it is that super-resolution imaging and original contrast imaging are compound
Stack result.
Specific implementation mode
The present invention is described in further detail with reference to the accompanying drawings and examples.
The present invention proposes a kind of basic for being designed through cranium dedicated cell domain volume imaging transducer arrays, and herein
On through cranial capacity super-resolution imaging method.
Referring to Fig. 1, through cranium dedicated cell domain volume imaging transducer using ultrasonic zonule two dimensional surface transducer array
Design：By N1×N2A array element composition, each array element size are l × l, and array element centre distance is d, in the battle array geometry of focal length face
The heart is F.Referring to Fig. 2, the volume imaging transducer is to use two-dimensional array to change through the special low frequency ultrasound volume energy converter of cranium
Energy device can collect the whole in single pass in a certain range using the scanning that can carry out both direction when energy converter
Then echo information carries out connection re-establishment 3-D view.
By taking following two-dimensional arrays as an example：It is 32 × 8 totally 256 array elements, main shaft through cranium dedicated cell domain volume imaging transducer
Array element center spacing be 1mm~1.4mm, main shaft effective aperture be 16mm~22.4mm, secondary axes array element center spacing be 1mm~
1.4mm, secondary axes effective aperture are 16mm~22.4mm, and centre frequency is 1.8MHz~2.2MHz, and the pulse width of energy converter is small
In equal to 10 μ s, bandwidth is more than or equal to 60%, and adjacent array element crosstalk is less than or equal to -30dB, and backing decaying is less than or equal to -60dB.
Meet actual design requirement, and diversified operating mode is provided.
It is used in the present invention through the special low frequency ultrasound volume energy converter of cranium, this energy converter can increase on minor axis direction
Imaging resolution and sensitivity are improved, and imaging can be scanned in two directions in effective aperture.Meanwhile main shaft has
It is smaller to imitate aperture, image areas that can be smaller in region can be good at being bonded temporal bone.The energy converter can be in the cell of temporal bone
The scanning of orthogonal both direction is carried out within the scope of domain, and then can be collected in a certain range once when being observed
Whole echo informations in scanning, realize real-time three-dimensional imaging.Its with bilateral scanning, in real time, low frequency, high-resolution and small
The features such as regional imaging.
Referring to Fig. 3, the observation bit point through cranium imaging of tissue is located at the most thin position of entire skull, i.e. Fig. 3 orbicular spots mark
Skull both sides temporal bone position, it can be seen that the region very little of temporal bone, and simultaneously out-of-flatness, thus at temporal bone carry out skull
Interior ultrasonic imaging can preferably be fitted in temporal bone to the more demanding of transducer dimensions, ultrasound volume energy converter of the invention
Position carry out imaging.
Super-resolution proposed by the present invention is through in cranial capacity imaging method：Volume energy converter emission space ultrasonic beam is direct
Obtain three-D ultrasonic volume data.Three-dimensional data is a large amount of continuous two obtained by the movement of volume transducer scans plane
Cross-section diagram is tieed up, then the two-dimensional image information of each section inputs computer after being digitized together with its location information.Often
One moment, the three-dimensional data information that the data that computer obtains are made of many sections, three-dimensional data information include a series of
Volume pixel, each pixel both include gray value, also include brightness value.Last computer carries out image to these three-dimensional datas
Reconstruction and data processing, can finally show on the screen.The key of volume imaging is that collected data
It rebuilds, then further carries out image procossing, imaging resolution is improved, to obtain the higher 3-D view of resolution ratio.This
It is special based on Markov Chain illiteracy for the voxel arest neighbors method (VNN) and three-dimensional of pixel to invent the 3-D view recovery algorithms being related to
The TCD,transcranial Doppler cerebral angiography super-resolution imaging method of Carlow multi-target tracking can be effectively improved through cranial capacity imaging point
Resolution.
Referring to Fig. 4, the cranium that making in laboratory is penetrated through the special low frequency ultrasound volume energy converter of cranium designed using the present invention
Bone model carries out microvesicle contrast imaging experiment to vascular pattern, since true skull is difficult to obtain and is not easy to operate,
Printed using 3D printing technique in actual experiment be closer to real human body tissue acoustic attenuation model (C.Bai, M.Ji,
J.Zong,et al.,“A3D-printed Skull Model with Corresponding Acoustic
Characteristic of Human Skull for Ultrasound Brain Imaging and Diagnosis”,
Proceedings of International Society for Therapeutic Ultrasound(ISTU)17th
Annual Symposium, Nanjing, China, Jun., 2017), model is interior to imitate body containing brain tissue, and contrast microbubbles are according to blood
Pipe imitates the pipeline of body and continues to flow through the imitative body of brain tissue with flow pumps promotion.Low frequency ultrasound transducer is in open ultrasonic imaging platform
Emit low frequency ultrasound signal, receives echo-signal under the control of host and echo-signal is sent to host；Host believes echo
Number it is transmitted to open ultrasonic imaging platform receiving module；Receiving module to receive echo signal sample, be stored as radio frequency and adopt
Sample data simultaneously send computer to；Computer is to echo radio frequency sampled data using at volume energy converter three-dimensional imaging algorithm
Reason, then shows to obtain ultrasonoscopy according to the imaging process of standard.
Referring to Fig. 5, volume energy converter three-dimensional imaging algorithm detailed process is as follows：
1) it is positioned near the temporal bone of side and (declines to the sound of ultrasonic wave in this region through cranium dedicated cell domain volume imaging transducer
Subtract most weak), emit plane of ultrasound wave to encephalic, obtains continuous N=100 frames ultrasonic contrast echo rf data, imaging frame rate Frt
For 1kHz；
2) two dimensional beam synthesis is carried out to echo rf data, obtains the data after all Beam synthesis；
3) according to the rf data after Beam synthesis, the Spatial Rules volume pixel network W of each frame imaging is determinedA×B×C,
Wherein, A is x-axis direction number of pixels, and B is z-axis direction number of pixels, and C is y-axis direction number of pixels.Regular voxel network
Structure needs to consider that two aspects, the size of the size of the size of voxel network and single tissue points, voxel network are required to
Surround all free scan images；The setting of single voxel volume size cannot be too small, too small to cause in reconstruction image " to divide
The appearance in boundary line ", while algorithm calculation amount is increased, but to rebuilding image error, there is no promoted with signal-to-noise ratio.Single body simultaneously
Pixel volume setting cannot be excessive, and it is apparent that the excessive single voxel of volume can cause reconstruction image to exist since voxel number is less
" sawtooth " phenomenon.The setting of voxel volume size should be such that voxel volume size approximation keeps to be divided into standard between Image Acquisition
Near the acquisition interval of scan image；
4) to each volume pixel point wa,b,m(a=1 ..., A；B=1 ..., B；M=1 ..., C), traverse each frame at
As all pixels point on all scan images, each pixel is calculated apart from Spatial Rules voxel network WA×B×CDistance, use
Apart from tissue points wa,b,cNearest pixel value is to the tissue points into row interpolation；
5) apart from tissue points wa,b,cNearest pixel value should be apart from tissue points wa,b,cIt should on the nearest plane of scanning motion
Tissue points wa,b,cProjection pixel w 'a,b,c.The threshold quantity h of addition distance is needed during realizingmaxIf apart from the body
The nearest pixel distance of vegetarian refreshments is more than hmax, then illustrate the tissue points farther out from free scan image, then need not use space freedom
Scan image can so reduce the time complexity of algorithm into row interpolation (i.e. the tissue points value is 0)；
To VNN algorithms, interpolation distance threshold hmaxDetermine whether the nearest pixel apart from tissue points can be used for this
Tissue points assignment, hmaxIt is too small, then it can lead to the presence of the blank voxel appearance that cannot be much interpolated in reconstruction image, influence three
Reconstructed results are tieed up, with hmaxIncrease, blank voxel gradually decreases to scanning area and is interpolated completely.H simultaneouslymaxBe arranged it is excessive,
Assignment may be then carried out to the region of certain separate scan images, noise is introduced in reconstructed results, so hmaxNumerical value determination
It is the h in the case where guarantee can be to voxel assignment in continuous scanning regionmaxIt is smaller；
6) the three-D volumes pixel network W ' that will be obtainedA×B×CIt is divided into N along y-axis1(N1=8, secondary axes array element number) two
Tie up image Bm(m=1 ..., N1), to the N frame images B of every piece imagem,n(n=1 ..., N) it chooses successively wherein continuously
NmedFrame obtains BM, %n(%n=n-Nmed+1..., n), to BM, %nIt is handled with time domain medium filtering function, obtains Background
As Gm,n：
Wherein, Distance () indicates that distance between pixels, i indicate that pixel abscissa, j indicate pixel ordinate；
7) from original image Bm,nIn subtract background image Gm,nObtain foreground image Fm,n, microvesicle and other movement targets
Image be included in foreground image；
8) dimensional Gaussian kernel function is built, by two-dimensional convolution to foreground image Fm,nFiltering, to reduce from non-microvesicle
The influence that the artefact of mobile target carrys out picture strip, obtains filtered foreground image F'm,n：
Wherein, naIndicate the lateral dimension of dimensional Gaussian kernel function, nbThe longitudinal size of expression dimensional Gaussian kernel function, two
Tieing up gaussian kernel function is：
Wherein, Δ z and Δ x is respectively the Pixel Dimensions of image axially and transversely,σxAnd σzBe by
The size of microvesicle in the picture determines；
9) to F'm,nGiven threshold fthr=-20dB, pixel value are less than fthrInformation removal, and carry out at local maximum
Reason；Then by N1Width foreground image synthesizes three-dimensional data according to former three-dimensional spatial distribution, then obtains each frame μnA pixel
As the position of contrast microbubbles, recording its coordinate isAnd can according to position spacing and when
Between difference obtain corresponding velocity information
10) all μ in n-th frame imagenThe combinatorial coordinates of a pixel, obtainAt all F ' m, n
Manage obtained YnCombination, obtains
11) set distance range in Y will be less than distance range centered on each target point (contrast microbubbles position)
Other target points be preset as centrales punctuate there may be the hyperlink target of track points；
12) referring to Fig. 6, for the three-dimensional TCD,transcranial Doppler cerebral angiography based on Markov chain Monte-Carlo multi-target tracking
Iterations L=2000 is deduced in different deduction states in super-resolution imaging procedure, initializing set path, is deduced
State includes eight kinds, respectively：New life, extinction, update, fracture, fusion, continuation, reduction and conversion.
13) track sum M under iterative process is calculated, if M=0, chooses newborn state (i.e. new life state probability 1)；If M
=1, then one kind is randomly selected in other deduction states in addition to transition status, i.e. each state probability 1/7；If M>1, then eight
One kind is randomly selected in kind of deduction state, i.e. each state probability 1/8；
14) according to the state of extraction, change the target point in Y, track collection ω=[ω is obtained to deduce1 ω2 …
ωM], wherein M is total number of tracks；
15) it setsWhereinIt indicates that track is concentrated to make
The coordinate information of shadow microvesicle,Indicate that the velocity information of contrast microbubbles is concentrated in track, k=1,2 ... K, K are
It is imaging frame number that contrast microbubbles number, n=1,2 ... N, N are concentrated in track, and footmark x and z indicate horizontal and vertical coordinate, structure respectively
Making time domain iterative state equation has：
Wherein A is process matrix：
Δ t=1/Frt, Frt are imaging frame rate,For the process variance matrix of Gaussian distributed；
16) according to Kalman filtering, the microbubble locations of prediction can be expressed as：
Wherein,Indicate observing matrix,For the observational variance square of Gaussian distributed
Battle array；If predicted position at this time is a false-alarm state, satisfaction is uniformly distributed
17) statistics calculates final on trajectory number zt, current track sum nt, newborn track starting point number at, it is selected as rail
The target points d of markt, false-alarm targets sum ft, and calculate from the new trace number c derived from last iterationtAnd it is not selected and is
The target points g of trackt：
ct=nt-zt-at
gt=nt-dt
18) Prior Probability under current deduction state is calculated：
Wherein, pzAnd pdIndicate that target termination or disappearance probability and target are selected the probability for tracing point, λ respectivelybAnd λf
The probability value and false-alarm probability value of target point new life in ideal trajectory are indicated respectively；
19) assume that false-alarm number and deduction path are obeyed and be uniformly distributed, calculate and deduce track collection and each rail based on current
The likelihood value of mark target point：
Wherein, M is to deduce total number of tracks, τkCurrently to deduce target point sum in track,For target
The likelihood value of point,WithObservation predicted vector and observational variance prediction matrix in Kalman filtering are indicated respectively；
20) posterior probability values under current deduction state are calculated：
21) it can be considered as certain markovian state currently to deduce track collection ω, according to Maximun Posterior Probability Estimation Method
It calculates and receives current deduction probability：
Wherein, ω ' is previous deduction track collection, and ω is this deduction track collection；
22) it generates and obeys equally distributed random number ξ (0<ξ<1), if ξ<A (ω ', ω) then receives this deduction, new rail
Mark collection ω will replace old track collection ω '；Otherwise, refuse this deduction, track collection ω ' is remained unchanged；
23) step 13) -22 is repeated), until completing iterations.
24) the final track collection of deduction is imaged.
On this basis, in conjunction with pulse inversion technique, coded excitation technology, microvesicle tracking super-resolution imaging technology, ginseng
Measure imaging technique etc., it can be achieved that different function through cranium blood vessel three-dimensional imaging.
It is two-dimensional super-resolution rate imaging results, 1 internal diameter 1mm of medium vessels, 2 internal diameter 0.7mm of blood vessel referring to Fig. 7, (a)；(b)
It is super-resolution imaging in (a) with original contrast imaging complex superposition as a result, being compared according to imaging results, shows using being based on horse
The TCD,transcranial Doppler cerebral angiography algorithm of Er Kefu chain Monte-Carlo multi-target trackings carries out encephalic super-resolution imaging and can be improved
Image resolution ratio.
TCD,transcranial Doppler cerebral angiography super-resolution imaging of the three-dimensional based on Markov chain Monte-Carlo multi-target tracking
Method be operation dimension is changed to three-dimensional on the basis of two-dimensional, basic principle remains unchanged, select VNN to scan image into
On the basis of row processing, Three-dimensional warp cranium brain may be implemented in the three-dimensional imaging method based on Markov chain Monte-Carlo multi-target tracking
The super-resolution imaging of blood vessel.
Claims (10)
1. a kind of ultrasound low frequency is through cranial capacity super-resolution three-dimensional contrast imaging method, it is characterised in that：Include the following steps：
1) two dimensional surface energy converter is utilized to emit plane of ultrasound wave to encephalic, the plane of ultrasound wave is in orthogonal two sides
It is scanned upwards, by receiving the echo information of the scanning, obtains continuous N frames ultrasonic contrast echo rf data；
2) two dimensional beam synthesis is carried out to the echo rf data, obtains the rf data B after Beam synthesisn(n=1 ...,
N)；
3) according to the rf data after Beam synthesis, the Spatial Rules volume pixel network W of each frame imaging is determinedA×B×C,
In, A is x-axis direction number of pixels, and B is z-axis direction number of pixels, and C is y-axis direction number of pixels；
4) to each volume pixel point wa,b,c(a=1 ..., A；B=1 ..., B；C=1 ..., C), it traverses in each frame imaging
All pixels point on all scan images calculates each pixel of scan image apart from the Spatial Rules volume pixel network
WA×B×CDistance, with apart from volume pixel point wa,b,cNearest pixel value is to volume pixel point wa,b,cInto row interpolation；Interpolation
After obtain and Spatial Rules volume pixel network WA×B×CThe identical three-D volumes pixel network W ' of volumeA×B×C；
5) the three-D volumes pixel network W ' that will be obtainedA×B×CIt is divided into N along y-axis1Width two dimensional image Bm(m=1 ..., N1), N1For
The two dimensional surface energy converter secondary axes array number, to Bm(m=1 ..., N1) in per piece image N frame images Bm,n(n=
1 ..., N) wherein continuous N is chosen successivelymedFrame obtains BM, %n(%n=n-Nmed+1..., n), to BM, %nIt is filtered with time domain intermediate value
Wave function is handled, and background image G is obtainedm,n：
Wherein, Distance () indicates that distance between pixels, i indicate that pixel abscissa, j indicate pixel ordinate；
6) from image Bm,nIn subtract background image Gm,nObtain foreground image Fm,n；
7) to foreground image Fm,nAfter carrying out artefact removal and local maximum processing, by N1Width foreground image is according to WA×B×CBody
Product and distribution splicing synthesize three-dimensional data, n-th frame image are obtained, then by μ in obtained each framenA pixel is used as and makes
The position of shadow microvesicle records the position coordinates of contrast microbubblesAnd corresponding velocity information
8) by all μ in n-th frame imagenThe combinatorial coordinates of a pixel, obtainBy all foreground images
Handle obtained YnCombination, obtains
9) set distance range in Y will be less than other contrast microbubbles of distance range centered on each contrast microbubbles position
There may be the link contrast microbubbles positions of track for contrast microbubbles position centered on position is default；
10) contrast microbubbles track is deduced using Markov chain Monte-Carlo multi-target tracking；
11) the final track collection of deduction is imaged.
2. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：It is 1mm~1.4mm, the effective hole of main shaft that the two dimensional surface energy converter, which uses 128 or 256 array elements, main shaft array element center spacing,
Diameter is 16mm~22.4mm, and secondary axes array element center spacing is 1mm~1.4mm, and secondary axes effective aperture is 16mm~22.4mm, center
Frequency is 1.8MHz~2.2MHz.
3. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：In the step 1), imaging frame rate Frt is 0.8kHz~1.2kHz.
4. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：The two dimensional beam synthesis uses the minimum variance adaptive beam composition algorithm in feature based space.
5. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：In the step 4), the nearest pixel value is apart from volume pixel point wa,b,cThe volume picture on the nearest plane of scanning motion
The pixel of the projection of vegetarian refreshments, if the volume pixel point is more than or equal to preset threshold value at a distance from the corresponding nearest plane of scanning motion
Amount does not then click through row interpolation to the volume pixel.
6. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：The artefact removal includes the following steps：
7.1) dimensional Gaussian kernel function is built, by two-dimensional convolution to foreground image Fm,nFiltering, obtains filtered foreground image
F′m,n：
Wherein, naIndicate the lateral dimension of dimensional Gaussian kernel function, nbIndicate that the longitudinal size of dimensional Gaussian kernel function, two dimension are high
This kernel function is：
Wherein, Δ z and Δ x is respectively the Pixel Dimensions of image axially and transversely,σxAnd σzIt is two-dimentional high
The variance of this kernel function；
7.2) given threshold fthr, remove F 'm,nIn be less than fthrPixel value；
The threshold value fthrFor -20dB~-15dB.
7. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 1
In：The step 10) specifically includes following steps：
10.1) according to the deduction state of extraction, change the contrast microbubbles position in Y, and deduce and obtain track collection ω=[ω1 ω2
… ωM], wherein M is total number of tracks；
10.2) time domain iterative state equation is constructed：
Wherein A is process matrix：
Δ t=1/Frt, Frt are imaging frame rate,For the process variance matrix of Gaussian distributed；
WhereinIndicate that the seat of contrast microbubbles is concentrated in track
Information is marked,It is that radiography is concentrated in track to indicate that the velocity information of contrast microbubbles, k=1,2 ... K, K are concentrated in track
Microvesicle number, n=1,2 ... N, N are imaging frame number, and footmark x and z indicate horizontal and vertical coordinate respectively；
10.3) according to Kalman filtering, the microbubble locations of prediction are expressed as：
Wherein, For the observational variance matrix of Gaussian distributed,To be uniformly distributed；
10.4) statistics calculates final on trajectory number zt, current track sum nt, newborn track starting point number at, it is selected as track
Target count dt, false-alarm targets sum ft, and calculate from the new trace number c derived from last iterationtAnd it is not selected as rail
The target points g of markt：
ct=nt-zt-at
gt=nt-dt
10.5) Prior Probability under current deduction state is calculated：
Wherein, pzAnd pdIndicate that target termination or disappearance probability and target are selected the probability for tracing point, λ respectivelybAnd λfRespectively
Indicate the probability value and false-alarm probability value of target point new life in ideal trajectory；
10.6) assume that false-alarm number and deduction path are obeyed and be uniformly distributed, calculate and deduce the track tracks Ji Hege based on current
The likelihood value of target point：
Wherein, M is to deduce total number of tracks, τkCurrently to deduce target point sum in track,For target point
Likelihood value,WithObservation predicted vector and observational variance prediction matrix in Kalman filtering are indicated respectively；
10.7) posterior probability values under current deduction state are calculated：
Wherein, P (Y) indicates the prior probability of detection contrast microbubbles position；
10.8) it currently to deduce track collection ω as certain markovian state, calculates and receives according to Maximun Posterior Probability Estimation Method
It is current to deduce probability：
Wherein, ω ' is previous deduction track collection, and ω is this deduction track collection；
10.9) it generates and obeys equally distributed random number ξ, 0<ξ<1, if ξ<A (ω ', ω), then receive this deduction, ω replaced
Change ω '；Otherwise, refuse this deduction, track collection remains unchanged；
10.10) step 10.1) -10.9 is repeated), until completing iterations.
8. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 7
In：The value of the N is 80~200, and it is 1500~3000 to deduce iterations L.
9. through cranial capacity super-resolution three-dimensional contrast imaging method, feature exists a kind of ultrasonic low frequency according to claim 7
In：The deduction state includes new life, extinction, update, fracture, fusion, continuation, reduction and conversion；The extraction packet of deduction state
Include following steps：It calculates track sum M under iterative process and chooses newborn state if M=0；If M=1, except conversion shape
One kind is randomly selected in other deduction states outside state；If M>1, then randomly select one kind in each deduction state.
10. a kind of ultrasound low frequency is through cranial capacity super-resolution three-dimensional angiographic imaging system, it is characterised in that：It is flat including two-dimensional ultrasound
Face energy converter and echo signal reception module, Cerebral vessels ultrasound contrast imaging module, contrast microbubbles extraction module, artefact removal
Module and super-resolution imaging module are deduced in module, contrast microbubbles track；Wherein：
The ultrasound echo signal that the echo signal reception module is used to receive the two-dimensional ultrasound planar transducer is adopted
Sample is simultaneously converted into rf data；
The Cerebral vessels ultrasound contrast imaging module is used for according to rf data, most using the voxel of two dimensional beam synthesis and pixel
Near neighbor method generates Cerebral vessels ultrasound contrastographic picture, and is converted into the form of Spatial Rules volume pixel network；
The contrast microbubbles extraction module is for removing the back of the body generated by time domain medium filtering in Cerebral vessels ultrasound contrastographic picture
Scape information, and retain the image information of contrast microbubbles；
The artefact removal module is used to filter out the motion track of the non-contrast microbubbles in the image for retaining contrast microbubbles；
It deduces module and is used for through Markov chain Monte-Carlo multi-target tracking to contrast microbubbles rail in the contrast microbubbles track
Mark is deduced；
The super-resolution imaging module is used to carry out cerebral angiography super-resolution imaging according to the track collection that deduction obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810201405.XA CN108324324A (en) | 2018-03-12 | 2018-03-12 | It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810201405.XA CN108324324A (en) | 2018-03-12 | 2018-03-12 | It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108324324A true CN108324324A (en) | 2018-07-27 |
Family
ID=62930719
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810201405.XA Pending CN108324324A (en) | 2018-03-12 | 2018-03-12 | It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108324324A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109513123A (en) * | 2018-12-28 | 2019-03-26 | 西安交通大学 | A kind of three-dimensional passive cavitation imaging method of the high-resolution based on half spherical array |
| CN110404085A (en) * | 2019-08-30 | 2019-11-05 | 浙江大学 | One kind wearing soft ultrasound gel material of cranium acoustics and its preparation method and application |
| CN111402210A (en) * | 2020-03-03 | 2020-07-10 | 山东师范大学 | Super-resolution positioning method and system for single-molecule fluorescence signal image |
| CN111588410A (en) * | 2020-05-18 | 2020-08-28 | 南京超维景生物科技有限公司 | Image processing method and device based on ultrasonic contrast image |
| CN111588409A (en) * | 2020-05-18 | 2020-08-28 | 南京超维景生物科技有限公司 | Super-resolution reconstruction method and device for three-dimensional ultrasonic contrast image |
| CN111784724A (en) * | 2020-05-28 | 2020-10-16 | 长安大学 | Improved Markov chain Monte Carlo two-dimensional rock slice reconstruction method and system |
| CN112888372A (en) * | 2018-10-19 | 2021-06-01 | 梅约医学教育与研究基金会 | System and method for kalman filter-based microvascular repair for super-resolution imaging |
| CN113499099A (en) * | 2021-07-21 | 2021-10-15 | 上海市同仁医院 | Carotid artery ultrasonic automatic scanning and plaque identification system and method |
| CN113724562A (en) * | 2021-07-20 | 2021-11-30 | 西安交通大学 | Simulated craniocerebral model for transcranial ultrasonic scanning and preparation method thereof |
| CN115601604A (en) * | 2022-11-29 | 2023-01-13 | 西南石油大学（Cn） | Multi-task microbubble track tracking method based on long-time and short-time memory network |
| CN116740265A (en) * | 2023-04-25 | 2023-09-12 | 华润武钢总医院 | Three-dimensional ultrasonic super-resolution rapid imaging method based on breast cancer diagnosis |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1973776A (en) * | 2005-11-28 | 2007-06-06 | 香港理工大学 | Three-dimensional ultrasonic detection method |
| CN101036586A (en) * | 2006-03-15 | 2007-09-19 | 株式会社东芝 | Ultrasound diagnostic apparatus and method for displaying ultrasound image |
| CN101958006A (en) * | 2010-09-03 | 2011-01-26 | 南京大学 | X-ray image-based three-dimensional object imaging method |
| CN103761767A (en) * | 2014-01-02 | 2014-04-30 | 华南理工大学 | Quick three-dimensional ultrasound image reconstruction method based on sparse data |
| CN104068845A (en) * | 2014-03-06 | 2014-10-01 | 武汉培威医学科技有限公司 | Electrocardio ultrasonic signal fusion computed tomography imaging system and method |
| CN104114099A (en) * | 2012-02-15 | 2014-10-22 | 株式会社日立制作所 | Ultrasonic imaging device |
| CN104323803A (en) * | 2014-10-30 | 2015-02-04 | 西安交通大学 | Vocal cord vibration imaging and measuring system and method on basis of plane wave ultrasonic imaging |
| CN104777485A (en) * | 2015-04-20 | 2015-07-15 | 西安交通大学 | Three-dimensional wide-beam small-region rapid cavitating and imaging method of ultrasonic two-dimensional planar array |
| CN104777484A (en) * | 2015-02-13 | 2015-07-15 | 西安交通大学 | Methods and systems for ultrasonic imaging and microbubble imaging of plane waves based on compressive adaptive beam forming |
| CN107073290A (en) * | 2014-10-24 | 2017-08-18 | 科莱希思有限公司 | Ultrasonic wave sleeve and utilize its ultrasonic therapeutic head |
| CN107753062A (en) * | 2017-11-27 | 2018-03-06 | 西安交通大学 | TCD,transcranial Doppler cerebral angiography super-resolution imaging method based on Markov chain Monte-Carlo multi-target tracking |
-
2018
- 2018-03-12 CN CN201810201405.XA patent/CN108324324A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1973776A (en) * | 2005-11-28 | 2007-06-06 | 香港理工大学 | Three-dimensional ultrasonic detection method |
| CN101036586A (en) * | 2006-03-15 | 2007-09-19 | 株式会社东芝 | Ultrasound diagnostic apparatus and method for displaying ultrasound image |
| CN101958006A (en) * | 2010-09-03 | 2011-01-26 | 南京大学 | X-ray image-based three-dimensional object imaging method |
| CN104114099A (en) * | 2012-02-15 | 2014-10-22 | 株式会社日立制作所 | Ultrasonic imaging device |
| CN103761767A (en) * | 2014-01-02 | 2014-04-30 | 华南理工大学 | Quick three-dimensional ultrasound image reconstruction method based on sparse data |
| CN104068845A (en) * | 2014-03-06 | 2014-10-01 | 武汉培威医学科技有限公司 | Electrocardio ultrasonic signal fusion computed tomography imaging system and method |
| CN107073290A (en) * | 2014-10-24 | 2017-08-18 | 科莱希思有限公司 | Ultrasonic wave sleeve and utilize its ultrasonic therapeutic head |
| CN104323803A (en) * | 2014-10-30 | 2015-02-04 | 西安交通大学 | Vocal cord vibration imaging and measuring system and method on basis of plane wave ultrasonic imaging |
| CN104777484A (en) * | 2015-02-13 | 2015-07-15 | 西安交通大学 | Methods and systems for ultrasonic imaging and microbubble imaging of plane waves based on compressive adaptive beam forming |
| CN104777485A (en) * | 2015-04-20 | 2015-07-15 | 西安交通大学 | Three-dimensional wide-beam small-region rapid cavitating and imaging method of ultrasonic two-dimensional planar array |
| CN107753062A (en) * | 2017-11-27 | 2018-03-06 | 西安交通大学 | TCD,transcranial Doppler cerebral angiography super-resolution imaging method based on Markov chain Monte-Carlo multi-target tracking |
Non-Patent Citations (1)
| Title |
|---|
| DIMITRI ACKERMANN D,SCHMITZ G: "Detection and Tracking of Multiple Microbubbles", 《IEEE TRANSACTION ON ULTRASONICS,FERROELECTRICS,AND FREQUENCY CONTROL》 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112888372A (en) * | 2018-10-19 | 2021-06-01 | 梅约医学教育与研究基金会 | System and method for kalman filter-based microvascular repair for super-resolution imaging |
| CN109513123B (en) * | 2018-12-28 | 2020-05-19 | 西安交通大学 | High-resolution three-dimensional passive cavitation imaging method based on hemispherical array |
| CN109513123A (en) * | 2018-12-28 | 2019-03-26 | 西安交通大学 | A kind of three-dimensional passive cavitation imaging method of the high-resolution based on half spherical array |
| CN110404085B (en) * | 2019-08-30 | 2020-09-22 | 浙江大学 | Transcranial acoustic soft ultrasonic gel material and preparation method and application thereof |
| CN110404085A (en) * | 2019-08-30 | 2019-11-05 | 浙江大学 | One kind wearing soft ultrasound gel material of cranium acoustics and its preparation method and application |
| CN111402210A (en) * | 2020-03-03 | 2020-07-10 | 山东师范大学 | Super-resolution positioning method and system for single-molecule fluorescence signal image |
| CN111588410A (en) * | 2020-05-18 | 2020-08-28 | 南京超维景生物科技有限公司 | Image processing method and device based on ultrasonic contrast image |
| CN111588409A (en) * | 2020-05-18 | 2020-08-28 | 南京超维景生物科技有限公司 | Super-resolution reconstruction method and device for three-dimensional ultrasonic contrast image |
| WO2021233188A1 (en) * | 2020-05-18 | 2021-11-25 | 南京超维景生物科技有限公司 | Super-resolution reconstruction method and apparatus for three-dimensional contrast-enhanced ultrasound image |
| WO2021233189A1 (en) * | 2020-05-18 | 2021-11-25 | 南京超维景生物科技有限公司 | Image processing method and apparatus based on ultrasound contrast image |
| CN111588409B (en) * | 2020-05-18 | 2023-04-07 | 南京超维景生物科技有限公司 | Super-resolution reconstruction method and device for three-dimensional ultrasonic contrast image |
| CN111784724A (en) * | 2020-05-28 | 2020-10-16 | 长安大学 | Improved Markov chain Monte Carlo two-dimensional rock slice reconstruction method and system |
| CN111784724B (en) * | 2020-05-28 | 2023-05-09 | 长安大学 | Improved Markov chain Monte Carlo two-dimensional rock slice reconstruction method and system |
| CN113724562A (en) * | 2021-07-20 | 2021-11-30 | 西安交通大学 | Simulated craniocerebral model for transcranial ultrasonic scanning and preparation method thereof |
| CN113499099A (en) * | 2021-07-21 | 2021-10-15 | 上海市同仁医院 | Carotid artery ultrasonic automatic scanning and plaque identification system and method |
| CN115601604A (en) * | 2022-11-29 | 2023-01-13 | 西南石油大学（Cn） | Multi-task microbubble track tracking method based on long-time and short-time memory network |
| CN115601604B (en) * | 2022-11-29 | 2023-04-07 | 西南石油大学 | Multi-task microbubble track tracking method based on long-time and short-time memory network |
| CN116740265A (en) * | 2023-04-25 | 2023-09-12 | 华润武钢总医院 | Three-dimensional ultrasonic super-resolution rapid imaging method based on breast cancer diagnosis |
| CN116740265B (en) * | 2023-04-25 | 2024-04-26 | 华润武钢总医院 | Three-dimensional ultrasonic super-resolution rapid imaging method based on breast cancer diagnosis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108324324A (en) | It is a kind of ultrasound low frequency through cranial capacity super-resolution three-dimensional contrast imaging method and system | |
| Huang et al. | A review on real-time 3D ultrasound imaging technology | |
| US11969286B2 (en) | Systems and methods for automatic detection and visualization of turbulent blood flow using vector flow data | |
| CN101917908B (en) | Quantification analisys of immobilized contrast agent in medical imaging applications | |
| JP4878251B2 (en) | Ultrasonic diagnostic apparatus and ultrasonic image display method | |
| EP1030191A2 (en) | Semi-automated segmentation method for 3-dimensional ultrasound | |
| CN108836392B (en) | Ultrasonic imaging method, device and equipment based on ultrasonic RF signal and storage medium | |
| KR20190038448A (en) | Measurement point determination in medical diagnostic imaging | |
| EP1722333B1 (en) | Method and device for reconstructing two-dimensional sectional images | |
| CN107753062A (en) | TCD,transcranial Doppler cerebral angiography super-resolution imaging method based on Markov chain Monte-Carlo multi-target tracking | |
| US20230329669A1 (en) | System and method for concurrent visualization and quantification of blood flow using ultrasound vector flow imaging | |
| CN110809801A (en) | System and method for simultaneous visualization and quantification of wall shear stress in blood vessels | |
| CN106030657B (en) | Motion Adaptive visualization in medicine 4D imaging | |
| US20100324420A1 (en) | Method and System for Imaging | |
| CN113678167A (en) | Segmentation and view guidance in ultrasound imaging and associated devices, systems, and methods | |
| CN113260314A (en) | System and method for contrast enhanced imaging | |
| WO2023039353A2 (en) | Real-time super-resolution ultrasound microvessel imaging and velocimetry | |
| CN110313943A (en) | Medical diagnostic apparatus, medical image-processing apparatus and image processing method | |
| McCall et al. | Non-invasive transcranial volumetric ultrasound localization microscopy of the rat brain with continuous, high volume-rate acquisition | |
| US7376254B2 (en) | Method for surface-contouring of a three-dimensional image | |
| CN114615937A (en) | System and method for vascular imaging | |
| Böhm et al. | The processing and analysis of radiographic image sequences | |
| Lengyel et al. | Three-dimensional reconstruction and volume rendering of intravascular ultrasound slices imaged on a curved arterial path | |
| US20220061803A1 (en) | Systems and methods for generating ultrasound probe guidance instructions | |
| Kanoulas et al. | Realistic Super-Resolution Image Analysis for State of the Art 2D Contrast Enhanced Ultrasound Imaging |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180727 |
|
| RJ01 | Rejection of invention patent application after publication |