CN109171811A - The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam - Google Patents
The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam Download PDFInfo
- Publication number
- CN109171811A CN109171811A CN201811115045.8A CN201811115045A CN109171811A CN 109171811 A CN109171811 A CN 109171811A CN 201811115045 A CN201811115045 A CN 201811115045A CN 109171811 A CN109171811 A CN 109171811A
- Authority
- CN
- China
- Prior art keywords
- cavitation
- frequency domain
- signal
- frequency
- steering vector
- 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
Classifications
-
- 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
- A61B8/4411—Device being modular
-
- 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
- A61B8/5207—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of raw data to produce diagnostic data, e.g. for generating an image
-
- 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
- A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Radiology & Medical Imaging (AREA)
- Biomedical Technology (AREA)
- Veterinary Medicine (AREA)
- Medical Informatics (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- General Health & Medical Sciences (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Ultra Sonic Daignosis Equipment (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The present invention provides a kind of passive cavitation imaging of frequency domain based on the synthesis of feature space adaptive beam and frequency multiplexed imaging methods, make Fourier transformation to time domain cavitation signal and carry out phase offset and array element apodization processing；Covariance matrix is constructed, the weight vectors of adaptive beam synthesis are obtained according to normalization steering vector and covariance matrix；Feature Space Decomposing is carried out to covariance matrix, weight vectors are projected to after signal subspace to the passive cavitation imaging for carrying out full frequency-domain；It constructs the passive cavitation imaging results of full frequency-domain that cavitation artifact is not known than index and according to different steering vectors under collection parameter and selects the uncertain collection parameter of optimal steering vector, to realize that the passive cavitation of different sub- frequency domains is imaged and frequency multiplexed is imaged.The present invention can significantly inhibit imaging artefacts, to improve the spatial discrimination performance of the passive cavitation imaging of frequency domain, can any number of frequency contents be carried out with comprehensive fixed sign, suitable for the real time monitoring of a variety of focused ultrasound therapies application.
Description
Technical field
The invention belongs to ultrasound detections and ultrasonic imaging technique field, and in particular to one kind is based on the adaptive wave of feature space
The passive cavitation imaging of the frequency domain of Shu Hecheng and frequency multiplexed imaging method.
Background technique
In ultrasound in medicine and biology field, ultrasonic cavitation is had been a hot spot of research, and in tumour heating ablation, external stone crushing, is surpassed
The focused ultrasound therapies such as sound thrombolysis and drug release cavitation has all served vital.It is empty in tumour heating ablation
Focused ultrasound beams formation can be blocked to influence focusing performance by changing microvesicle, caused damage field unpredictable, reduced treatment
Accuracy；And the sound thermal conversion efficiency of tissue can be improved in another aspect cavitation, thus when reducing treatment energy and shortening treatment
Between.In rubble, the energy that discharges generates in regional area when cavitation microvesicle moment collapses high temperature, high pressure, shock wave and
The physical phenomenons such as jet stream accelerate the broken of calculus.In ultrasound thrombolysis, the high rapid fire of cavitation vibrating microbubbles and rupture generation
Stream and shock wave etc. can effectively decompose thrombus, to improve thrombolysis efficiency.However due to the uncontrollability of cavitation and at random
Property, may normal tissue form irreversible damage, therapeutic purposes are not achieved.Even in addition, studies have shown that examining
Under disconnected ultrasound acoustic pressure, the shearing force and jet stream that cavitation microvesicle is formed can also generate biological effect to peripheral cell or tissue, thus
Generate the risk of injury of blood vessel and rupture.Therefore efficiently control and using cavitation for focused ultrasound therapy scheme regulation with
Optimize important in inhibiting, and to realize this point and then have to the development dependent on cavitation detection technique.
In recent years, cavitation detection technique has developed to two dimension from one-dimensional, utilization clinically common linear array and phased array etc.
Ultrasonic diagnostic transducer, which can be realized, determines sign to cavitation spatial distribution.Traditional B-mode ultrasonic imaging, plane wave have been developed at present
Supper-fast ultrasonic imaging and the ultrasound gradually actives such as scanning imagery cavitation imaging method based on timing control, pass through and adjust hair
Emission mode improves imaging algorithm performance simultaneously can realize high contrast and the imaging of high-resolution cavitation.However, due to focusing
Active cavitation can be imaged during ultrasonic therapy and generate strong interference, therefore can only select to stop making completely in focusing ultrasound
With being imaged later or before focusing ultrasonic pulse next time and not arriving also, and just due to this point, the imaging of active cavitation
The cavitation microvesicle for focusing and generating when ultrasound does not act on can only be detected, and cannot detect and focus what ultrasonication generated in the process
Cavitation behavior cannot also realize the real time monitoring imaging of focused ultrasound therapy.In order to make up the deficiency of active cavitation imaging, closely
There is scholar to propose a kind of passive cavitation imaging method over year, this method only passively receives cavitation signal due to not emitting,
Not by the interference of focus ultrasound signals, so as to realize that the effective of focused ultrasound therapy process cavitation determines sign.Passive cavitation
Imaging has been used for monitoring tissue heating ablation, tissue ablation, drug release and Blood Brain Barrier (BBB) opening etc..It is most popular at present
And most common passive cavitation imaging method is to carry out in the time domain, have later scholar propose the passive cavitation of frequency domain at
Picture.Compared to time domain passive cavitation imaging, the passive cavitation imaging of frequency domain can directly frequency domain selection imaging object be stable cavitation or
It is inertial cavitation, the band information only selected does Beam synthesis, therefore can reduce calculation amount compared to time domain approach；In addition,
The imaging of frequency domain passive cavitation directly by the time delays of Fourier transformation realization signal, can to avoid due to low sampling rate from
Dissipate delay time error caused by time sampling.However the passive cavitation imaging method of traditional frequency domain is that one kind is disobeyed intrinsically
Rely the blind beam synthesizing method in data self character, will cause extremely serious imaging artefacts, image quality is bad, Bu Nengti
Altitude sterically defined precision, therefore it is unfavorable for the real time monitoring of focused ultrasound therapy process.Therefore how high score is realized
The passive cavitation imaging of the frequency domain of resolution is a major challenge of this field, and needs overcome difficult point.
Summary of the invention
The purpose of the present invention is to provide a kind of passive cavitation imagings of frequency domain based on the synthesis of feature space adaptive beam
And frequency multiplexed imaging method.
To achieve the goals above, the invention adopts the following technical scheme:
A kind of passive cavitation imaging method of frequency domain based on the synthesis of feature space adaptive beam, comprising the following steps:
Step 1: passively receiving cavitation radiofrequency signal by ultrasonic imaging energy converter, carries out Fourier to cavitation radiofrequency signal
Transformation obtains frequency domain cavitation signal, utilizes normalization cosine apodization function progress after making phase offset processing to frequency domain cavitation signal
Apodization processing obtains the frequency domain cavitation letter after phase offset and array element apodization processing of all array elements of ultrasonic imaging energy converter
Number；
Step 2: empty according to the frequency domain after phase offset and array element apodization processing of the resulting all array elements of step 1
Change signal and construct covariance matrix, is synthesized using the covariance matrix and the uncertain collection parameter building adaptive beam of steering vector
Cost function, Lagrange's multiplier is introduced by constraint condition to the cost function and constructs Ge Lang constraint function, it is right
Lagrangian constraint function derivation and to enable derived function be zero, obtains steering vector expression formula, steering vector expression formula is substituted into institute
It states in the constraint condition of cost function and solves to obtain Lagrange's multiplier using Newton iteration method, the glug obtained according to solution
Bright day multiplier calculates steering vector and steering vector is simultaneously made normalized, according to steering vector after normalized and described
Covariance matrix obtains the weight vectors of adaptive beam synthesis；
Step 3: carrying out Feature Space Decomposing to the covariance matrix, obtain signal subspace, and step 2 is resulting
The weight vectors of adaptive beam synthesis project on signal subspace, resulting to step 1 using the weight vectors after projection
The frequency domain cavitation signal after phase offset and array element apodization processing of all array elements is weighted summation, then removes direct current
Component obtains the output signal of feature space adaptive beam synthesis；To the output signal of feature space adaptive beam synthesis
Progress square is simultaneously overlapped on entire frequency domain, obtains the passive cavitation imaging results of full frequency-domain；
Step 4: it according to cavitation artifact than selecting optimal steering vector is uncertain to collect parameter, is sweared according to optimal guiding
The optimal output signal of the uncertain collection parameter determining features spatially adaptive Beam synthesis of amount；Exist respectively to the optimal output signal
It is overlapped on the sub- frequency domain of difference in entire frequency domain, obtains the passive cavitation imaging results of different sub- frequency domains.
A kind of passive cavitation frequency multiplexed imaging method based on the synthesis of feature space adaptive beam, including following step
It is rapid:
Step 1: Step 2: step 3 and step 4 are identical as the passive cavitation imaging method of above-mentioned frequency domain；
Step 5: standardization is done to the passive cavitation imaging results of the sub- frequency domain of difference, then by two of them
The above sub- frequency domain is overlapped by the passive cavitation imaging results of standardization, obtains passive cavitation frequency multiplexed imaging
As a result.
Above-mentioned steps one, specifically includes the following steps:
1.1) time domain cavitation signal is passively received using ultrasonic imaging energy converter, the time domain cavitation signal refers to that focusing is super
The cavitation radiofrequency signal scattered at sonic transducer focal regions；
1.2) Fourier transformation is carried out to the time domain cavitation signal that i-th of array element of ultrasonic imaging energy converter receives, obtained
Frequency domain cavitation signal:
Wherein, i=1,2 ..., N, N are ultrasonic imaging transducer array element number, t1It is the beginning for acquiring time domain cavitation signal
Moment, t2It is the stop timing for acquiring time domain cavitation signal, j is imaginary unit, pi(t, x, z) be i-th of array element receive when
Domain cavitation signal, f indicate frequency；
1.3) phase offset processing is carried out to frequency domain cavitation signal obtained by step 1.2):
Wherein, pofi(f, x, z) is the phase offset of i-th of array element；
1.4) array element apodization processing is carried out to the resulting phase offset of step 1.3) treated frequency domain cavitation signal:
Wherein, Apodi(f, x, z) is the normalization cosine apodization function of i-th of array element；
1.5) step 1.2)~1.4 are repeated), the process phase offset of N number of array element until obtaining ultrasonic imaging energy converter
With the frequency domain cavitation signal after array element apodization processing.
Above-mentioned steps two, specifically includes the following steps:
2.1) the frequency domain cavitation after phase offset and array element apodization processing of the resulting all array elements of step 1 is utilized
Signal constructs covariance matrix R:
Wherein, S (f, x, z)=[S1(f,x,z)；S2(f,x,z)；...；SN(f, x, z)] be ultrasonic imaging energy converter N
The frequency domain cavitation signal matrix after phase offset and array element apodization processing of a array element,It indicates for arbitrary frequency f；
2.2) cost function of the building adaptive beam synthesis of covariance matrix obtained by step 2.1) is utilized:
Wherein, a is the steering vector to be solved,For the steering vector of hypothesis, | | | | Euclid norm is represented,For constraint condition, ε is the uncertain collection parameter of steering vector；
2.3) it after step 2.2), introduces Lagrange's multiplier and constructs Lagrangian constraint function:
Wherein, λ is Lagrange's multiplier；
2.4) the resulting Lagrangian constraint function of step 2.3) to steering vector a derivation and is enabled into derived function obtained by derivation
It is zero, then utilizes matrix inversion lemma, obtain steering vector expression formula:
Wherein, I is unit matrix；
2.5) the resulting steering vector expression formula of step 2.4) is substituted into the constraint item of the resulting cost function of step 2.2)
In part, then constraint condition is rewritten are as follows:
2.6) by the Feature Space Decomposing of the resulting covariance matrix R of step 2.1) to the resulting constraint item of step 2.5)
Part is converted, then the lagrangian multiplier in constraint condition after conversion is then solved using Newton iteration method is passed through
The resulting steering vector expression formula of step 2.4) calculates steering vector, and does normalized, obtains normalization steering vector
2.7) it is calculated certainly according to the resulting normalization steering vector of step 2.6) and the resulting covariance matrix R of step 2.1)
Adapt to the weight vectors w of Beam synthesis:
Above-mentioned steps three, specifically includes the following steps:
3.1) empty to the frequency domain after phase offset and array element apodization processing of all array elements obtained according to step 1
Change covariance matrix constructed by signal and carry out Feature Space Decomposing:
R=U Λ UH=RS+RI
Wherein, U=[u1,u2,...,uN] it is characterized vector matrix, Λ=diag [γ1,γ2,...,γN] it is to angular moment
Battle array, the matrix diagonals element are characterized value, and γ1≥γ2≥...≥γN, RSFor the covariance matrix of signal subspace, RIFor
The covariance matrix of noise subspace；
3.2) weight vectors that adaptive beam synthesizes are projected into signal subspace, the weight vectors after being projected
Wherein, To normalize steering vector, US=[u1,u2,...,uL] it is preceding L larger characteristic values
The signal subspace that corresponding feature vector is constituted, ΛS=diag [γ1,γ2,...,γL] be signal subspace characteristic value
Matrix, diag [] indicate the diagonalization of vector；L is characterized the number of all characteristic values in value greater than δ times of maximum eigenvalue,
δ is 0.2~0.5；
3.3) empty using the frequency domain after phase offset and array element apodization processing of each array element of ultrasonic imaging energy converter
Weight vectors after changing signal and the resulting projection of step 3.2) calculate the weighted frequency domain cavitation signal of each array element:
Wherein,For the weight vectors after the resulting projection of step 3.2)I-th of element, Si(f, x, z) is ultrasound
The frequency domain cavitation signal after phase offset and array element apodization processing of i-th of array element of imaging transducer, wherein i=1,
2 ..., N, N are ultrasonic imaging transducer array element number；
3.4) the weighted frequency domain cavitation signal of the N number of array element of ultrasonic imaging energy converter is overlapped, obtain signal Q (f, x,
Z):
3.5) DC component is got rid of from signal Q (f, x, z) obtained by step 3.4), then obtains the adaptive wave of feature space
The output signal of Shu HechengIt is rightCarry out square simultaneously be overlapped on entire frequency domain, obtain it is each at
The energy of image position (x, z):
Wherein,It indicates for arbitrary frequency f.
Above-mentioned steps four, specifically includes the following steps:
4.1) cavitation artifact of the construction based on average energy is than index CAR:
CAR=20log10(ImeanCav/ImeanArt)
Wherein, ImeanCavFor cavitation zone or the average energy of cavitation zone selected part, ImeanArtFor artifact region or puppet
The average energy of shadow zone domain selected part, the unit of CAR are dB；
4.2) portion is difference selected with artifact region according to cavitation zone and the cavitation artifact ratio or cavitation zone of artifact region
The mean value of the cavitation artifact ratio divided, evaluates the passive cavitation image quality of full frequency-domain under different ε, and ε is the uncertain collection of steering vector
Parameter, and choose the corresponding ε when CAR or CAR mean value highest and do not know to collect parameter as optimal steering vector, according to optimal
The uncertain collection parameter of steering vector obtain the optimal output signal of feature space adaptive beam synthesisIts
In, the traversal range of ε is 0.01~N, and traversal step-length is 0.01~1；
4.3) different sub- frequency domains is selected, the optimal output to the synthesis of step 4.2) resulting feature space adaptive beam
Signal is overlapped respectively in different sub- frequency domains, then obtains the passive cavitation imaging of different sub- frequency domains:
Wherein, FR indicates given a certain sub- frequency domain, f1For the lower-frequency limit of the sub- frequency domain, f2For the frequency of the sub- frequency domain
The upper limit.
Above-mentioned steps five, specifically includes the following steps:
5.1) the passive cavitation imaging results of the sub- frequency domain of difference resulting to step 4 are standardized, and are then carried out
Superposition, obtains passive cavitation frequency multiplexed imaging results:
Wherein, NFR is the sub- frequency domain number for participating in frequency multiplexed imaging,To participate in frequency multiplexed imaging
The passive cavitation imaging results of i-th of sub- frequency domain after standardization.
A kind of passive cavitation imaging system of frequency domain based on the synthesis of feature space adaptive beam, including phase offset and battle array
First apodization processing module, adaptive beam synthesize the passive cavitation image-forming module and son of weight vectors computing module, full frequency-domain
The passive cavitation image-forming module of frequency domain；The phase offset and array element apodization processing module are for executing above-mentioned steps one；It is described
Adaptive beam synthesis weight vectors computing module is for executing above-mentioned steps two；The passive cavitation image-forming module of the full frequency-domain
For executing above-mentioned steps three；The passive cavitation image-forming module of the sub- frequency domain is for executing above-mentioned steps four.
A kind of passive cavitation frequency multiplexed imaging system based on the synthesis of feature space adaptive beam, including above-mentioned phase
Offset and array element apodization processing module, above-mentioned adaptive beam synthesize the passive sky of weight vectors computing module, above-mentioned full frequency-domain
Change the passive cavitation image-forming module and passive cavitation frequency multiplexed image-forming module of image-forming module, above-mentioned sub- frequency domain；It is described passive
Cavitation frequency multiplexed image-forming module is for executing above-mentioned steps five.
The beneficial effects of the present invention are embodied in:
The present invention is directed to the passive cavitation imaging method of traditional frequency domain existing defect in terms of monitoring focused ultrasound therapy, leads to
It crosses and phase offset and array element apodization processing is carried out to frequency domain cavitation signal and construct the covariance matrix of frequency domain, building frequency domain is certainly
The cost function for adapting to Beam synthesis acquires the weight vectors of adaptive beam synthesis on this basis；Then by weight vectors
The covariance matrix of frequency domain is projected to through the weight vectors on the resulting signal subspace of Feature Space Decomposing, and after use projection
Beam synthesis is carried out, effective inhibition has been carried out to imaging artefacts, improves the space point of the different passive cavitation imagings of sub- frequency domain
It distinguishes performance, has great significance for the real time monitoring of focused ultrasound therapy process cavitation.
The present invention is based on the passive cavitation imagings of the sub- frequency domain of high-resolution performance, by the passive cavitation imaging results of sub- frequency domain
It is standardized and superposition realizes passive cavitation frequency multiplexed imaging, further improve image quality.
Detailed description of the invention
Fig. 1 is that time domain cavitation signal is transformed into frequency domain and carries out phase offset to frequency domain cavitation signal in the embodiment of the present invention
With the flow chart of array element apodization；
Fig. 2 is the time domain cavitation signal that certain ultrasonic linear-array energy converter is got and process phase offset and array element apodization processing
Frequency domain cavitation signal afterwards；
Fig. 3 is the calculation flow chart of the weight vectors of adaptive beam synthesis in the embodiment of the present invention；
Fig. 4 is covariance matrix, Lagrange's multiplier iterative value, steering vector and weight vectors in adaptive beam synthesis
Result；
Fig. 5 is the passive cavitation Irnaging procedures of full frequency-domain based on the synthesis of feature space adaptive beam in the embodiment of the present invention
Figure；
Fig. 6 is characterized the covariance matrix of signal and noise subspace in spatially adaptive Beam synthesis, adding after projection
The result of weight vector and the passive cavitation imaging of full frequency-domain；
Fig. 7 is the stream of the imaging of passive cavitation and the passive cavitation frequency multiplexed imaging of different sub- frequency domains in the embodiment of the present invention
Cheng Tu；
Fig. 8 is the imaging of passive cavitation and passive cavitation frequency multiplexed imaging results of different sub- frequency domains.
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawings and examples.
The present invention provides a kind of imaging of frequency domain passive cavitation and frequency multiplexed based on the synthesis of feature space adaptive beam
Imaging method, comprising the following steps:
Step 1: ultrasonic imaging energy converter passively receives time domain cavitation signal, carries out Fourier's change to time domain cavitation signal
It gets frequency domain cavitation signal in return, and phase offset processing is made to frequency domain cavitation signal, then utilize normalization cosine apodization function
Obtain the frequency domain cavitation signal after phase offset and array element apodization processing.Specific step is as follows (Fig. 1):
(1.1) the only passive mode for receiving signal is set by supersonic imaging apparatus, is write by arbitrary waveform generator
Waveform triggers focused transducer and supersonic imaging apparatus to synchronize, passive by the ultrasonic imaging energy converter of supersonic imaging apparatus
The cavitation radiofrequency signal scattered at focal regions in collectiong focusing ultrasonic transducer mechanism, is known as time domain for the cavitation radiofrequency signal
Cavitation signal, the time domain cavitation signal are acquired by the parallel channel data acquisition of supersonic imaging apparatus and memory module；
(1.2) assume that ultrasonic imaging energy converter (for example, ultrasonic linear-array energy converter) has N number of array element, i-th of array element is acquired
The time domain cavitation signal arrived carries out Fourier transformation, obtains frequency domain cavitation signal Pi(f, x, z):
Wherein i=1,2 ..., N, t1At the beginning of being acquisition time domain cavitation signal, t2It is acquisition time domain cavitation signal
Stop timing, j are imaginary unit, pi(t, x, z) is the collected time domain cavitation signal of i-th of array element, and t indicates the time；
(1.3) it calculates from imaging position (x, z) to i-th of array element (xi, 0) ultrasonic transmission time, i.e. i-th array element when
Between be delayed:
Wherein, c is the acoustic propagation velocity of ultrasonic wave；
(1.4) phase offset of i-th of array element is calculated according to the resulting time delays of step (1.3):
Wherein, τiFor τiThe abbreviation of (x, z)；
(1.5) to obtain phase according to frequency domain cavitation signal obtained by step (1.2) and step (1.4) resulting phase offset inclined
Move treated frequency domain cavitation signal
(1.6) normalization apodizing function is calculated according to frequency domain cavitation signal obtained by step (1.2):
Wherein,It indicates for arbitrary frequency f；
(1.7) according to the position (x of imaging position (x, z) and i-th of array elementi, 0) and calculate cosine apodization function:
(1.8) returned according to cosine apodization function calculating obtained by normalization apodizing function obtained by step (1.6) and step (1.7)
One changes cosine apodization function:
Apodi(f, x, z)=Apod1i(f,x,z)·Apod2i(f,x,z)
(1.9) using normalization cosine apodization function obtained by step (1.8) to the processing of step (1.5) resulting phase offset
Frequency domain cavitation signal afterwards carries out array element apodization processing:
(1.10) next array element is jumped to, step (1.2)~(1.9) are repeated, until obtaining the process phase of N number of array element
Frequency domain cavitation signal after offset and array element apodization processing.
Referring to fig. 2, wherein (a) is ultrasonic imaging transducer array schematic diagram, array element sum is 128, and pore size is
38mm, (b) is the 64th collected time domain cavitation signal of array element, (c) is the process phase offset and array element of the 64th array element
Frequency domain cavitation signal after apodization processing.
Step 2: empty according to the frequency domain after phase offset and array element apodization processing of the resulting all array elements of step 1
Change signal to construct covariance matrix, is closed using the covariance matrix and the uncertain collection parameter building adaptive beam of steering vector
At cost function, Lagrange's multiplier is introduced to the constraint condition of cost function and obtains steering vector expression formula, utilizes ox
Iterative method of pausing acquires the Lagrange's multiplier in constraint condition, calculates steering vector on this basis and makees at normalization to it
Reason finally obtains the weight vectors of adaptive beam synthesis.Specific step is as follows (Fig. 3):
(2.1) the frequency domain cavitation after phase offset and array element apodization processing of the resulting N number of array element of step 1 is utilized
Signal constructs covariance matrix:
Wherein, S (f, x, z)=[S1(f,x,z)；S2(f,x,z)；...；SN(f, x, z)] be N number of array element process phase
Frequency domain cavitation signal matrix after offset and array element apodization processing, []HRepresent the conjugate transposition of matrix；
(2.2) cost function of the building adaptive beam synthesis of covariance matrix obtained by step (2.1) is utilized:
Wherein, a is the steering vector to be solved,For the steering vector of hypothesis, | | | | Euclid norm is represented,For constraint condition, ε is the parameter for describing the uncertain collection of steering vector, referred to as uncertain collection parameter, []-1Generation
Table matrix is inverted；
(2.3) it for cost function described in solution procedure (2.2), introduces Lagrange's multiplier and constructs Lagrange about
Beam function:
Wherein, λ is Lagrange's multiplier；
(2.4) to steering vector a derivation and enable derivation is resulting to lead Lagrangian constraint function described in step (2.3)
Function is zero, then obtains steering vector expression formula:
Wherein, R is step (2.1) resulting covariance matrix R (f, x, z), and hereinafter abbreviated as R, I are unit matrix；
(2.5) steering vector expression formula described in step (2.4) is rewritten according to matrix inversion lemma are as follows:
(2.6) the resulting steering vector expression formula of step (2.5) is substituted into the constraint condition of step (2.2)
In, then constraint condition is rewritten are as follows:
(2.7) it calculates for convenience, Feature Space Decomposing R=U Λ U is done to covariance matrix RH, wherein U is feature vector
Matrix, Λ is characterized value matrix, and enablesThen step (2.6) resulting constraint condition can convert are as follows:
Wherein, biFor i-th of element of b, γiIt is characterized i-th of diagonal element of value matrix Λ；
(2.8) Lagrange's multiplier in the constraint condition after step (2.7) conversion is solved using Newton iteration method
λ will solve obtained λ and substitute into step (2.5) resulting steering vector expression formula, then steering vector can be calculated, to meter
The steering vector calculated does normalized:
In Newton iteration method described in step (2.8), the condition of convergence and iteration initial value are set first, when the condition of convergence is full
When sufficient, iteration stopping obtains Lagrange's multiplier.The iterative relation formula of Newton iteration method are as follows:
Wherein k=1,2 ..., λkFor the kth time iterative value of Lagrange's multiplier, λk+1For the kth+1 of Lagrange's multiplier
Secondary iterative value, f (λk) be kth time iteration function, f'(λk) it is f (λk) derivative；Lagrange's multiplier iteration initial value λ1If
It is set toWherein γ1For the 1st diagonal element of step (2.7) the eigenvalue matrix Λ；
(2.9) it is calculated according to the resulting normalization steering vector of step (2.8) and the resulting covariance matrix of step (2.1)
The weight vectors of adaptive beam synthesis:
Referring to fig. 4, wherein (a) is empty according to the frequency domain after phase offset and array element apodization processing of 128 array elements
Change the covariance matrix that signal obtains；(b) change curve that Lagrange's multiplier increases with the number of iterations, it can be seen that the 6th time
Start to restrain when iteration；(c) it is normalized steering vector, is (d) weight vectors of adaptive beam synthesis.
Step 3: carrying out Feature Space Decomposing to covariance matrix obtained by step 2, obtains signal subspace, and by step
The weight vectors of two resulting adaptive beam synthesis project on signal subspace, using the weight vectors after projection to step
The frequency domain cavitation signal after phase offset and array element apodization processing of one resulting all array elements is weighted summation, then goes
DC component is removed, the output signal of feature space adaptive beam synthesis is obtained；To the synthesis of feature space adaptive beam
Output signal progress square is simultaneously overlapped on entire frequency domain, obtains the passive cavitation imaging results of full frequency-domain.Specific steps
As follows (Fig. 5):
(3.1) performance of adaptive beam composition algorithm declines mainly due to caused by the disturbance of noise subspace,
Middle signal subspace includes useful cavitation information, and noise subspace then includes the pseudo- shadow information for influencing image quality.For
Artifact is effectively inhibited, covariance matrix R resulting to step (2.1) carries out Feature Space Decomposing:
R=U Λ UH=RS+RI
Wherein, U=[u1,u2,...,uN] it is characterized vector matrix, Λ=diag [γ1,γ2,...,γN] it is to angular moment
Battle array, diagonal element are characterized value, γ1≥γ2≥...≥γN, RSFor the covariance matrix of signal subspace, RIIt is empty for noise
Between covariance matrix；
The covariance matrix of step (3.1) resulting signal subspace specifically:
Wherein, US=[u1,u2,...,uL] it is the feature vector square that the corresponding feature vector of the larger characteristic value of preceding L is constituted
Battle array, ΛS=diag [γ1,γ2,...,γL] be signal subspace eigenvalue matrix, diag [] indicate vector it is diagonal
Change；
The covariance matrix of step (3.1) resulting noise subspace specifically:
Wherein, UI=[uL+1,uL+2,...,uN] be the feature that constitutes of the corresponding feature vector of remaining smaller characteristic value to
Moment matrix, ΛI=diag [γL+1,γL+2,...,γN] be noise subspace eigenvalue matrix；
Parameter L is characterized value γ1,γ2,...,γNIn be greater than maximum eigenvalue (γ1) δ times all characteristic values number
Mesh, δ are typically chosen in 0.2~0.5；
(3.2) it enablesThen the weight vectors of the resulting adaptive beam synthesis of step (2.9) may be expressed as:
Wherein,For the resulting normalization steering vector of step (2.8)；
Since ideally, signal subspace is orthogonal to noise subspace, therefore the weighting that adaptive beam is synthesized
Weight vectors of the vector project to signal subspace, after can must being projected according to step (3.5) are as follows:
(3.3) the frequency domain cavitation after phase offset and array element apodization processing of the resulting each array element of step 1 is utilized
Weight vectors after signal and step (3.2) resulting projection calculate the weighted frequency domain cavitation signal of each array element:
Wherein,For the weight vectors after projection obtained by step (3.2)I-th of element, Si(f, x, z) is step 1
The frequency domain cavitation signal after phase offset and array element apodization processing of resulting i-th of array element；
(3.4) the weighted frequency domain cavitation signal of the N number of array element of ultrasonic imaging energy converter is overlapped, then can be obtained:
(3.5) DC component is got rid of from signal Q (f, x, z) obtained by step (3.4), then it is adaptive obtains feature space
The output signal of Beam synthesis:
(3.6) to the output signal progress square of step (3.5) resulting feature space adaptive beam synthesis and whole
It is overlapped on a frequency domain, obtains the energy of each imaging position (x, z), the i.e. passive cavitation imaging results of full frequency-domain:
Referring to Fig. 6, wherein (a) is covariance matrix of signal subspace, (b) covariance matrix of noise subspace, (c)
Weight vectors for adaptive beam synthesis project to the weight vectors obtained after signal subspace, (d) passively empty for full frequency-domain
It is melted into the result of picture.
Step 4: construction cavitation artifact selects optimal steering vector according to cavitation artifact than index than calculated result
Uncertain collection parameter, so that it is determined that the optimal output signal of feature space adaptive beam synthesis；Son frequency is obtained on this basis
The passive cavitation imaging results in domain are done standardization to the passive cavitation imaging results of different sub- frequency domains and are overlapped, obtain
To passive cavitation frequency multiplexed imaging results.Specific step is as follows (Fig. 7):
(4.1) cavitation artifact of the construction based on average energy is than index:
CAR=20log10(ImeanCav/ImeanArt)
Wherein ImeanCavFor the average energy of cavitation zone, ImeanArtUnit for the average energy of artifact region, CAR is
dB；
(4.2) traversal 0.01~N of range of the uncertain collection parameter ε of setting, traversal step-length generally may be configured as 0.01~1,
According to Step 1: method described in two and three, obtains the passive cavitation imaging knot of full frequency-domain under different uncertain collection parameter ε
Fruit；
(4.3) it is calculated under different uncertain collection parameter ε according to cavitation artifact described in step (4.1) than index CAR
The CAR of the passive cavitation imaging results of full frequency-domain is imaged with this to evaluate the passive cavitation of full frequency-domain under different uncertain collection parameter ε
Quality, and choose when CAR highest corresponding ε as optimal uncertain collection parameter；
The boundary of cavitation zone and artifact region is according to Energy maximum value in the passive cavitation imaging results of full frequency-domain in transverse direction
It is determined with lateral coordinates and axial coordinate corresponding when axially dropping to Energy maximum value half, is cavitation area in boundary
Domain is artifact region outside boundary；
(4.4) it carries out on the basis of optimal uncertain collection parameter ε Step 1: step (3.1) in step 2 and step 3
Processing described in~(3.5) obtains the optimal output signal of feature space adaptive beam synthesis
(4.5) different sub- frequency domains is selected in entire frequency domain, obtains different frequency contents, such as subharmonic (f0/2±
f0/ 8), harmonic wave (nf0±f0/ 4, n=2,3 ..., 10), ultraharmonics [(2n+1) f0/2±f0/ 8, n=1,2 ..., 10] and it is wide
Band noise [nf0/2+f0/ 8~(n+1) f0/2-f0/ 8, n=1,2 ..., 20] etc., wherein f0It is changed to generate the aggregation ultrasound of cavitation
The tranmitting frequency of energy device；
(4.6) to the optimal output signal of the synthesis of feature space adaptive beam described in step (4.4) in step (4.5)
The sub- frequency domain of difference is overlapped, then the passive cavitation imaging of different sub- frequency domains can be obtained:
Wherein, f1For the lower-frequency limit of sub- frequency domain, f2For the upper frequency limit of sub- frequency domain, FR indicates a given sub- frequency domain；
(4.7) following standardization is carried out to the passive cavitation imaging results of the resulting sub- frequency domain of difference of step (4.6):
Wherein,
Indicate that, for arbitrary imaging position (x, z), M is the total number of imaging position；
(4.8) the passive cavitation imaging results of the sub- frequency domain of difference after step (4.7) resulting standardization are carried out
Superposition, then can be obtained passive cavitation frequency multiplexed imaging results:
Wherein, NFR is the sub- frequency domain number for participating in frequency multiplexed imaging；
Referring to Fig. 8, wherein (a), (b), (c), (d) is according to the passive obtained differences of cavitation imaging method of traditional frequency domain
The imaging results of sub- frequency domain (subharmonic, harmonic wave, ultraharmonics, broadband noise)；(e), (f), (g), (h) are to be proposed according to the present invention
Based on feature space adaptive beam synthesis the obtained sub- frequency domain of difference of the passive cavitation imaging method of frequency domain (subharmonic,
Harmonic wave, ultraharmonics, broadband noise) imaging results；Compare (a)~(d) and (e)~(h) it is found that method proposed by the present invention can
Effectively inhibit the artifact of the passive cavitation imaging of traditional frequency domain, thus improves the spatial positioning accuracy of cavitation.In addition, Fig. 8 (i) is
Passive cavitation frequency multiplexed imaging results, through overfrequency (above-mentioned subharmonic, harmonic wave, ultraharmonics, four kinds of frequencies of broadband noise at
Point) it is compound after, it can be seen that relative to the imaging results of each sub- frequency domain, image quality is further increased.
For Fig. 8's (i) as a result, the present invention proposes following analysis: during focusing ultrasonication, preceding cavitation scattering
Multiple frequency contents between interfere with each other, and then influence subsequent cavitation behavior, and a needle is imaged in the passive cavitation of traditional frequency domain
A certain sub- frequency domain is imaged, cavitation activity is cannot achieve in the effective of multiple sub- frequency domains and determines sign, while this is also to cause
One reason of image quality decline.
The invention has the following advantages that
(1) collected cavitation scattered signal is carried out Beam synthesis in frequency domain by the present invention, directly can select frequency in frequency domain
Rate ingredient, effectively improves computational efficiency；Beam synthesis uses the phase offset of frequency domain in the process and the time of non-temporal prolongs
When, thus reduce as practical supersonic imaging apparatus sample rate is limited generate delay time error and caused by imaging artefacts；
(2) method proposed by the present invention is to carry out adaptive beam synthesis in frequency domain, solves the passive cavitation of traditional frequency domain
The problem of imaging method is independent of data self character, and weight vectors are projected on signal subspace, to side
Valve artifact remains cavitation information while significantly inhibition, thus realize the passive cavitation of high-resolution frequency domain at
Picture；
(3) present invention has selected optimal uncertain collection parameter than this index using cavitation artifact, on this basis
The high-resolution for having obtained the corresponding sub- frequency domain of the different frequencies ingredient such as subharmonic, harmonic wave, ultraharmonics and broadband noise is passively empty
It is melted into picture, so that the real-time detection and quantitative analysis for stable cavitation and inertial cavitation during focusing ultrasonication have established base
Plinth；
(4) present invention carries out the passive cavitation imaging of different sub- frequency domains compound, has obtained high-resolution frequency multiplexed
Imaging, image quality further increase, and the frequency multiplexed imaging method proposed can be to different sub- frequency domain width and different son frequency
Multiple frequency contents under the number of domain carry out comprehensive fixed sign；
(5) the passive cavitation imaging of frequency domain proposed by the present invention and frequency multiplexed imaging method can be to different focusing ultrasonic transductions
Device, different ultrasonic imaging energy converters, the different cavitations generated under ultrasound emission parameter and different experiments setting that focus are imaged,
It is suitable for different therapentic parts and different focused ultrasound therapy applications simultaneously, furthermore can also be used for ultrasonic cavitation transient state physics mistake
The analysis and research of journey.
Claims (9)
1. it is a kind of based on feature space adaptive beam synthesis the passive cavitation imaging method of frequency domain, it is characterised in that: including with
Lower step:
Step 1: passively receiving cavitation radiofrequency signal by ultrasonic imaging energy converter, carries out Fourier transformation to cavitation radiofrequency signal
Frequency domain cavitation signal is obtained, utilizes normalization cosine apodization function progress apodization after making phase offset processing to frequency domain cavitation signal
Processing, obtains the frequency domain cavitation signal after phase offset and array element apodization processing of all array elements of ultrasonic imaging energy converter；
Step 2: believed according to the frequency domain cavitation after phase offset and array element apodization processing of the resulting all array elements of step 1
Number construction covariance matrix, utilizes the generation of the uncertain collection parameter building adaptive beam synthesis of the covariance matrix and steering vector
Valence function introduces Lagrange's multiplier by the constraint condition to the cost function and constructs Ge Lang constraint function, to glug
Bright day constraint function derivation and to enable derived function be zero, obtains steering vector expression formula, steering vector expression formula is substituted into the generation
It solves in the constraint condition of valence function and using Newton iteration method to obtain Lagrange's multiplier, the Lagrange obtained according to solution
Multiplier calculate steering vector simultaneously steering vector is made into normalized, according to after normalized steering vector and the association side
Poor matrix obtains the weight vectors of adaptive beam synthesis；
Step 3: carrying out Feature Space Decomposing to the covariance matrix, obtain signal subspace, and step 2 is resulting adaptive
The weight vectors of Beam synthesis are answered to project on signal subspace, it is resulting to step 1 all using the weight vectors after projection
The frequency domain cavitation signal after phase offset and array element apodization processing of array element is weighted summation, then removes direct current point
Amount obtains the output signal of feature space adaptive beam synthesis；To feature space adaptive beam synthesis output signal into
Row square is simultaneously overlapped on entire frequency domain, obtains the passive cavitation imaging results of full frequency-domain；
Step 4: according to cavitation artifact than selecting optimal steering vector is uncertain to collect parameter, not according to optimal steering vector
Determine the optimal output signal of collection parameter determining features spatially adaptive Beam synthesis；To the optimal output signal respectively entire
It is overlapped on the sub- frequency domain of difference in frequency domain, obtains the passive cavitation imaging results of different sub- frequency domains.
2. a kind of passive cavitation frequency multiplexed imaging method based on the synthesis of feature space adaptive beam, it is characterised in that: packet
Include following steps:
Step 1: passively receiving cavitation radiofrequency signal by ultrasonic imaging energy converter, carries out Fourier transformation to cavitation radiofrequency signal
Frequency domain cavitation signal is obtained, utilizes normalization cosine apodization function progress apodization after making phase offset processing to frequency domain cavitation signal
Processing, obtains the frequency domain cavitation signal after phase offset and array element apodization processing of all array elements of ultrasonic imaging energy converter；
Step 2: believed according to the frequency domain cavitation after phase offset and array element apodization processing of the resulting all array elements of step 1
Number construction covariance matrix, utilizes the generation of the uncertain collection parameter building adaptive beam synthesis of the covariance matrix and steering vector
Valence function introduces Lagrange's multiplier by the constraint condition to the cost function and constructs Ge Lang constraint function, to glug
Bright day constraint function derivation and to enable derived function be zero, obtains steering vector expression formula, steering vector expression formula is substituted into the generation
It solves in the constraint condition of valence function and using Newton iteration method to obtain Lagrange's multiplier, the Lagrange obtained according to solution
Multiplier calculate steering vector simultaneously steering vector is made into normalized, according to after normalized steering vector and the association side
Poor matrix obtains the weight vectors of adaptive beam synthesis；
Step 3: carrying out Feature Space Decomposing to the covariance matrix, obtain signal subspace, and step 2 is resulting adaptive
The weight vectors of Beam synthesis are answered to project on signal subspace, it is resulting to step 1 all using the weight vectors after projection
The frequency domain cavitation signal after phase offset and array element apodization processing of array element is weighted summation, then removes direct current point
Amount obtains the output signal of feature space adaptive beam synthesis；To feature space adaptive beam synthesis output signal into
Row square is simultaneously overlapped on entire frequency domain, obtains the passive cavitation imaging results of full frequency-domain；
Step 4: according to cavitation artifact than selecting optimal steering vector is uncertain to collect parameter, not according to optimal steering vector
Determine the optimal output signal of collection parameter determining features spatially adaptive Beam synthesis；To the optimal output signal respectively entire
It is overlapped on the sub- frequency domain of difference in frequency domain, obtains the passive cavitation imaging results of different sub- frequency domains；
Step 5: doing standardization to the passive cavitation imaging results of the sub- frequency domain of difference, then will be more than two of them
Sub- frequency domain is overlapped by the passive cavitation imaging results of standardization, obtains passive cavitation frequency multiplexed imaging knot
Fruit.
3. imaging method according to claim 1 or 2, it is characterised in that: the step 1 specifically includes the following steps:
1.1) time domain cavitation signal is passively received using ultrasonic imaging energy converter, the time domain cavitation signal refers to that focusing ultrasound changes
The cavitation radiofrequency signal scattered at energy device focal regions；
1.2) Fourier transformation is carried out to the time domain cavitation signal that i-th of array element of ultrasonic imaging energy converter receives, obtains frequency domain
Cavitation signal:
Wherein, i=1,2 ..., N, N are ultrasonic imaging transducer array element number, t1At the beginning of being acquisition time domain cavitation signal,
t2It is the stop timing for acquiring time domain cavitation signal, j is imaginary unit, pi(t, x, z) is that the time domain that i-th of array element receives is empty
Change signal, f indicates frequency；
1.3) phase offset processing is carried out to frequency domain cavitation signal obtained by step 1.2):
Wherein, pofi(f, x, z) is the phase offset of i-th of array element；
1.4) array element apodization processing is carried out to the resulting phase offset of step 1.3) treated frequency domain cavitation signal:
Wherein, Apodi(f, x, z) is the normalization cosine apodization function of i-th of array element；
1.5) step 1.2)~1.4 are repeated), until obtaining the process phase offset and battle array of N number of array element of ultrasonic imaging energy converter
Frequency domain cavitation signal after first apodization processing.
4. imaging method according to claim 1 or 2, it is characterised in that: the step 2 specifically includes the following steps:
2.1) the frequency domain cavitation signal after phase offset and array element apodization processing of the resulting all array elements of step 1 is utilized
Construct covariance matrix R:
Wherein, S (f, x, z)=[S1(f,x,z)；S2(f,x,z)；...；SN(f, x, z)] be ultrasonic imaging energy converter N number of battle array
The frequency domain cavitation signal matrix after phase offset and array element apodization processing of member,It indicates for arbitrary frequency f；
2.2) cost function of the building adaptive beam synthesis of covariance matrix obtained by step 2.1) is utilized:
Wherein, a is the steering vector to be solved,For the steering vector of hypothesis, | | | | Euclid norm is represented,
For constraint condition, ε is the uncertain collection parameter of steering vector；
2.3) it after step 2.2), introduces Lagrange's multiplier and constructs Lagrangian constraint function:
Wherein, λ is Lagrange's multiplier；
2.4) to steering vector a derivation and the derived function obtained by derivation is enabled to be the resulting Lagrangian constraint function of step 2.3)
Zero, matrix inversion lemma is then utilized, steering vector expression formula is obtained:
Wherein, I is unit matrix；
2.5) the resulting steering vector expression formula of step 2.4) is substituted into the constraint condition of the resulting cost function of step 2.2),
Then constraint condition is rewritten are as follows:
2.6) by the Feature Space Decomposing of the resulting covariance matrix R of step 2.1) to the resulting constraint condition of step 2.5) into
Then row conversion is solved the lagrangian multiplier in the constraint condition after conversion using Newton iteration method, then passes through step
2.4) resulting steering vector expression formula calculates steering vector, and does normalized, obtains normalization steering vector
2.7) according to the resulting normalization steering vector of step 2.6)It is calculated with the resulting covariance matrix R of step 2.1) adaptive
Answer the weight vectors w of Beam synthesis:
5. imaging method according to claim 1 or 2, it is characterised in that: the step 3 specifically includes the following steps:
3.1) the frequency domain cavitation after phase offset and array element apodization processing of all array elements obtained according to step 1 is believed
Covariance matrix constructed by number carries out Feature Space Decomposing:
R=U Λ UH=RS+RI
Wherein, U=[u1,u2,...,uN] it is characterized vector matrix, Λ=diag [γ1,γ2,...,γN] it is diagonal matrix, it should
Matrix diagonals element is characterized value, and γ1≥γ2≥...≥γN, RSFor the covariance matrix of signal subspace, RIFor noise
The covariance matrix in space；
3.2) weight vectors that adaptive beam synthesizes are projected into signal subspace, the weight vectors after being projected
Wherein, To normalize steering vector, US=[u1,u2,...,uL] it is that preceding L larger characteristic values are corresponding
Feature vector constitute signal subspace, ΛS=diag [γ1,γ2,...,γL] be signal subspace eigenvalue matrix,
The diagonalization of diag [] expression vector；L is characterized the number of all characteristic values in value greater than δ times of maximum eigenvalue, and δ is
0.2~0.5；
3.3) believed using the frequency domain cavitation after phase offset and array element apodization processing of each array element of ultrasonic imaging energy converter
Number and the resulting projection of step 3.2) after weight vectors calculate the weighted frequency domain cavitation signal of each array element:
Wherein,For the weight vectors after the resulting projection of step 3.2)I-th of element, Si(f, x, z) changes for ultrasonic imaging
The frequency domain cavitation signal after phase offset and array element apodization processing of energy i-th of array element of device, wherein i=1,2 ..., N, N
For ultrasonic imaging transducer array element number；
3.4) the weighted frequency domain cavitation signal of the N number of array element of ultrasonic imaging energy converter is overlapped, obtains signal Q (f, x, z):
3.5) DC component is got rid of from signal Q (f, x, z) obtained by step 3.4), then obtains the conjunction of feature space adaptive beam
At output signalIt is rightProgress square is simultaneously overlapped on entire frequency domain, is obtained each at image position
Set the energy of (x, z):
Wherein,It indicates for arbitrary frequency f.
6. imaging method according to claim 1 or 2, it is characterised in that: the step 4 specifically includes the following steps:
4.1) cavitation artifact of the construction based on average energy is than index CAR:
CAR=20log10(ImeanCav/ImeanArt)
Wherein, ImeanCavFor cavitation zone or the average energy of cavitation zone selected part, ImeanArtFor artifact region or pseudo- shadow zone
The average energy of domain selected part；
4.2) according to the cavitation artifact ratio or cavitation zone and artifact region difference selected part of cavitation zone and artifact region
The mean value of cavitation artifact ratio evaluates the passive cavitation image quality of full frequency-domain under different ε, and ε is the uncertain collection parameter of steering vector,
And it chooses the corresponding ε when CAR or CAR mean value highest and is led as the uncertain collection parameter of optimal steering vector according to optimal
The optimal output signal that feature space adaptive beam synthesizes is obtained to the uncertain collection parameter of vectorWherein, ε
Traversal range is 0.01~N, and traversal step-length is 0.01~1；
4.3) different sub- frequency domains is selected, to the optimal output signal of step 4.2) resulting feature space adaptive beam synthesis
It is overlapped respectively in different sub- frequency domains, then obtains the passive cavitation imaging of different sub- frequency domains:
Wherein, FR indicates given a certain sub- frequency domain, f1For the lower-frequency limit of the sub- frequency domain, f2For the upper frequency limit of the sub- frequency domain.
7. imaging method according to claim 2, it is characterised in that: the step 5 specifically includes the following steps:
5.1) the passive cavitation imaging results of the sub- frequency domain of difference resulting to step 4 are standardized, and are then folded
Add, obtain passive cavitation frequency multiplexed imaging results:
Wherein, NFR is the sub- frequency domain number for participating in frequency multiplexed imaging,For the process for participating in frequency multiplexed imaging
The passive cavitation imaging results of the sub- frequency domain of i-th of standardization.
8. a kind of passive cavitation imaging system of frequency domain based on the synthesis of feature space adaptive beam, it is characterised in that: including phase
The passive cavitation imaging of position offset and array element apodization processing module, adaptive beam synthesis weight vectors computing module, full frequency-domain
The passive cavitation image-forming module of module and sub- frequency domain；The phase offset and array element apodization processing module are used for ultrasonic imaging
Passively received cavitation radiofrequency signal is fourier transformed and is converted to frequency domain cavitation signal energy converter, and to frequency domain cavitation signal
Make to carry out apodization processing using normalization cosine apodization function after phase offset processing, so that it is all to obtain ultrasonic imaging energy converter
The frequency domain cavitation signal after phase offset and array element apodization processing of array element；The adaptive beam synthesizes weight vectors meter
It calculates the frequency domain cavitation signal that module is used for according to all array elements after phase offset and array element apodization and constructs covariance square
Battle array, using the uncertain collection parameter building adaptive beam synthesis of the covariance matrix and steering vector cost function, by pair
The constraint condition of the cost function introduces Lagrange's multiplier construction Ge Lang constraint function, will be by constraining Lagrange
Function derivation and to enable derived function be that zero resulting steering vector expression formula is substituted into the constraint condition of the cost function and utilized
Newton iteration method solves Lagrange's multiplier, calculates steering vector and by steering vector according to solving obtained Lagrange's multiplier
Make normalized, and according to through normalized steering vector and the covariance matrix obtain adaptive beam synthesis
Weight vectors；The weight vectors that the passive cavitation image-forming module of the full frequency-domain is used to synthesize the adaptive beam project
To institute on using the resulting signal subspace of Feature Space Decomposing of the covariance matrix, using the weight vectors after projection
The frequency domain cavitation signal after phase offset and array element apodization processing for stating all array elements is weighted summation, removal direct current point
It is overlapped on entire frequency domain after amount, and square processing, to obtain the passive cavitation imaging results of full frequency-domain；The son
The passive cavitation image-forming module of frequency domain is used for according to cavitation artifact than selecting optimal steering vector is uncertain to collect parameter, Yi Jigen
According to the optimal output signal of the uncertain collection parameter determining features spatially adaptive Beam synthesis of optimal steering vector, and to this
Optimal output signal is overlapped on the sub- frequency domains of difference in entire frequency domain respectively, to obtain the passive sky of different sub- frequency domains
Change imaging results.
9. a kind of passive cavitation frequency multiplexed imaging system based on the synthesis of feature space adaptive beam, it is characterised in that: packet
It includes phase offset and array element apodization processing module, adaptive beam synthesizes the passive cavitation of weight vectors computing module, full frequency-domain
The passive cavitation image-forming module and passive cavitation frequency multiplexed image-forming module of image-forming module, sub- frequency domain；The phase offset and
Array element apodization processing module is used for that passively received cavitation radiofrequency signal to be fourier transformed and is converted to ultrasonic imaging energy converter
Frequency domain cavitation signal, and normalization cosine apodization function progress apodization is utilized after making phase offset processing to frequency domain cavitation signal
Processing, to obtain the frequency domain cavitation letter after phase offset and array element apodization processing of all array elements of ultrasonic imaging energy converter
Number；The adaptive beam synthesis weight vectors computing module is used to be become according to all array elements by phase offset and array element
Frequency domain cavitation signal after mark is constructed covariance matrix, is constructed certainly using the covariance matrix and the uncertain collection parameter of steering vector
It adapts to the cost function of Beam synthesis, introduce Lagrange's multiplier construction Ge Lang by the constraint condition to the cost function
Constraint function, will be by Lagrangian constraint function derivation and to enable derived function be that zero resulting steering vector expression formula substitutes into institute
The Lagrange for stating in the constraint condition of cost function and solving Lagrange's multiplier using Newton iteration method, being obtained according to solution
Multiplier calculates steering vector and steering vector is made normalized, and according to steering vector through normalized and described
Covariance matrix obtains the weight vectors of adaptive beam synthesis；The passive cavitation image-forming module of the full frequency-domain is used for will be described
The weight vectors of adaptive beam synthesis project to empty using the resulting signal subspace of Feature Space Decomposing of the covariance matrix
Between it is upper, empty using frequency domain by phase offset and array element apodization processing after of the weight vectors after projection to all array elements
Change after signal is weighted summation, removal DC component, and square processing and be overlapped on entire frequency domain, to obtain complete
The passive cavitation imaging results of frequency domain；The passive cavitation image-forming module of the sub- frequency domain is used for more optimal than selecting according to cavitation artifact
The uncertain collection parameter of steering vector, and according to the uncertain collection parameter determining features spatially adaptive wave of optimal steering vector
The optimal output signal of Shu Hecheng, and to being folded in the optimal the output signal respectively sub- frequency domains of difference in entire frequency domain
Add, to obtain the passive cavitation imaging results of different sub- frequency domains；The passive cavitation frequency multiplexed image-forming module is used for institute
The passive cavitation imaging results for stating different sub- frequency domains do standardization, and by the process standard of more than two of them sub- frequency domain
The passive cavitation imaging results for changing processing are overlapped, to obtain passive cavitation frequency multiplexed imaging results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811115045.8A CN109171811B (en) | 2018-09-25 | 2018-09-25 | The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811115045.8A CN109171811B (en) | 2018-09-25 | 2018-09-25 | The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109171811A true CN109171811A (en) | 2019-01-11 |
| CN109171811B CN109171811B (en) | 2019-08-23 |
Family
ID=64909806
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811115045.8A Active CN109171811B (en) | 2018-09-25 | 2018-09-25 | The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109171811B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110801267A (en) * | 2019-10-31 | 2020-02-18 | 西安交通大学 | Ultrasonic fine efficient thrombolytic system assisted by low-intensity focusing vortex sound field |
| CN111134719A (en) * | 2019-12-19 | 2020-05-12 | 西安交通大学 | Active and passive ultrasonic composite imaging method and system for phase-change nano liquid drops through focused ultrasonic irradiation |
| CN111265245A (en) * | 2020-01-20 | 2020-06-12 | 西安交通大学 | Passive cavitation imaging method and system based on double-constraint robust Capon beam synthesis and multiple apodization cross-correlation |
| CN113820728A (en) * | 2020-06-19 | 2021-12-21 | 河南省科学院应用物理研究所有限公司 | Beam forming interference suppression method and device for searching uncertain set |
| CN114519752A (en) * | 2021-12-31 | 2022-05-20 | 西安交通大学 | High-resolution fast-calculation passive ultrasonic imaging method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050154306A1 (en) * | 2004-01-14 | 2005-07-14 | Michael Burcher | Dort process-based method and system for adaptive beamforming in estimating the aberration in a medium |
| CN105223567A (en) * | 2015-09-28 | 2016-01-06 | 中国科学院声学研究所 | A kind of robust wideband Adaptive beamformer method being applied to ultrasonic imaging |
| CN106842212A (en) * | 2017-03-31 | 2017-06-13 | 西安交通大学 | The quick self-adapted beam synthesizing method of multiple apodization of feature based spatial decomposition |
| CN107260217A (en) * | 2017-07-17 | 2017-10-20 | 西安交通大学 | The three-dimensional passive imaging method and system monitored in real time for brain focused ultrasonic cavitation |
-
2018
- 2018-09-25 CN CN201811115045.8A patent/CN109171811B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050154306A1 (en) * | 2004-01-14 | 2005-07-14 | Michael Burcher | Dort process-based method and system for adaptive beamforming in estimating the aberration in a medium |
| CN105223567A (en) * | 2015-09-28 | 2016-01-06 | 中国科学院声学研究所 | A kind of robust wideband Adaptive beamformer method being applied to ultrasonic imaging |
| CN106842212A (en) * | 2017-03-31 | 2017-06-13 | 西安交通大学 | The quick self-adapted beam synthesizing method of multiple apodization of feature based spatial decomposition |
| CN107260217A (en) * | 2017-07-17 | 2017-10-20 | 西安交通大学 | The three-dimensional passive imaging method and system monitored in real time for brain focused ultrasonic cavitation |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110801267A (en) * | 2019-10-31 | 2020-02-18 | 西安交通大学 | Ultrasonic fine efficient thrombolytic system assisted by low-intensity focusing vortex sound field |
| CN111134719A (en) * | 2019-12-19 | 2020-05-12 | 西安交通大学 | Active and passive ultrasonic composite imaging method and system for phase-change nano liquid drops through focused ultrasonic irradiation |
| CN111134719B (en) * | 2019-12-19 | 2021-01-19 | 西安交通大学 | Active and passive ultrasonic composite imaging method and system for phase-change nano liquid drops through focused ultrasonic irradiation |
| CN111265245A (en) * | 2020-01-20 | 2020-06-12 | 西安交通大学 | Passive cavitation imaging method and system based on double-constraint robust Capon beam synthesis and multiple apodization cross-correlation |
| CN111265245B (en) * | 2020-01-20 | 2021-02-09 | 西安交通大学 | Passive cavitation imaging method and system based on double-constraint robust Capon beam synthesis and multiple apodization cross-correlation |
| CN113820728A (en) * | 2020-06-19 | 2021-12-21 | 河南省科学院应用物理研究所有限公司 | Beam forming interference suppression method and device for searching uncertain set |
| CN113820728B (en) * | 2020-06-19 | 2024-05-10 | 河南省科学院应用物理研究所有限公司 | Beam forming interference suppression method and device for uncertain set search |
| CN114519752A (en) * | 2021-12-31 | 2022-05-20 | 西安交通大学 | High-resolution fast-calculation passive ultrasonic imaging method and system |
| CN114519752B (en) * | 2021-12-31 | 2024-03-29 | 西安交通大学 | High-resolution rapid-calculation passive ultrasonic imaging method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109171811B (en) | 2019-08-23 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109171811B (en) | The passive cavitation imaging of frequency domain and frequency multiplexed imaging method based on the synthesis of feature space adaptive beam | |
| CN105722462B (en) | Relevant spread spectrum in synthetic aperture image is formed | |
| US11826585B2 (en) | Adaptive refocusing of ultrasound transducer arrays using image data | |
| Nilsen et al. | Beamspace adaptive beamforming for ultrasound imaging | |
| CN104169739B (en) | Spread spectrum coding waveform in ultrasonic imaging | |
| CN102946945B (en) | Efficient ultrasonic focuses on | |
| CN107260217B (en) | The passive imaging method of three-dimensional and system for brain focused ultrasonic cavitation real time monitoring | |
| US7273455B2 (en) | Corrections for wavefront aberrations in ultrasound imaging | |
| JPH04244145A (en) | Method for converting sonar image | |
| CN105520752B (en) | Beam forming device, supersonic imaging apparatus and beam-forming method | |
| CN102895000A (en) | Double-focusing beamforming method based on self-adaptive weighting | |
| Lu et al. | Passive acoustic mapping of cavitation using eigenspace-based robust Capon beamformer in ultrasound therapy | |
| Mehdizadeh et al. | Minimum variance beamforming applied to ultrasound imaging with a partially shaded aperture | |
| CN111265245B (en) | Passive cavitation imaging method and system based on double-constraint robust Capon beam synthesis and multiple apodization cross-correlation | |
| CN109513123A (en) | A kind of three-dimensional passive cavitation imaging method of the high-resolution based on half spherical array | |
| Hemmsen et al. | Tissue harmonic synthetic aperture ultrasound imaging | |
| CN105741236B (en) | Ultrasound system image broad sense disappears secondary lobe method | |
| Bera et al. | Multiline 3D beamforming using micro-beamformed datasets for pediatric transesophageal echocardiography | |
| KR20140102994A (en) | A method, apparatus and HIFU system for generating ultrasound forming multi-focuses in region of interest | |
| Lai et al. | Modeling of the acoustic field produced by diagnostic ultrasound arrays in plane and diverging wave modes | |
| Lu et al. | Passive cavitation mapping using dual apodization with cross-correlation in ultrasound therapy monitoring | |
| KR101911470B1 (en) | Apparatus and method for generating 3D ultrasonic image using plane wave | |
| Jeong et al. | High-spatial-resolution, instantaneous passive cavitation imaging with temporal resolution in histotripsy: a simulation study | |
| WO2017093778A1 (en) | An imaging method, an apparatus implementing said method, a computer program and a computer-readable storage medium | |
| Sivadon et al. | 3D passive cavitation imaging using adaptive beamforming and matrix array transducer with random apodization |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |