CN111466949A - MMSE (minimum mean square error) beam former, MMSE beam forming method and computer-readable storage medium - Google Patents
MMSE (minimum mean square error) beam former, MMSE beam forming method and computer-readable storage medium Download PDFInfo
- Publication number
- CN111466949A CN111466949A CN202010288748.1A CN202010288748A CN111466949A CN 111466949 A CN111466949 A CN 111466949A CN 202010288748 A CN202010288748 A CN 202010288748A CN 111466949 A CN111466949 A CN 111466949A
- Authority
- CN
- China
- Prior art keywords
- beamformer
- mmse
- mvdr
- echo data
- dcr
- 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
Images
Classifications
-
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
- A61B8/48—Diagnostic techniques
-
- 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
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Radiology & Medical Imaging (AREA)
- Heart & Thoracic Surgery (AREA)
- Biophysics (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Biomedical Technology (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Molecular Biology (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)
Abstract
The invention provides an MMSE (minimum mean square error) beam former, an MMSE beam forming method and a computer readable storage medium, which improve the technical scheme of realizing the MMSE (minimum mean square error) criterion of synthesized beam approximation based on a Minimum Variance Distortionless Response (MVDR) beam former and a scalar filter in the prior art, and rebuild a coherent factor calculation model and a covariance estimation model. The method solves the technical problem of low quality of the synthesized image caused by covariance estimation and coherence factor calculation deviation in the existing Minimum Variance Distortionless Response (MVDR) beam forming method, can adaptively inhibit the influence of data noise and interference on the imaging contrast ratio, and improves the quality of the synthesized image.
Description
Technical Field
The invention relates to the technical field of coherent plane compounding, in particular to an MMSE (minimum mean square error) beam former, an MMSE beam forming method and a computer-readable storage medium.
Background
Coherent plane wave synthesis (CPWC), a method to achieve high frame rate and uniform resolution throughout the imaging region, can use plane wave transmission to generate the entire image within one pulse-echo sequence, unlike conventional systems that generate images row by row.
The Minimum Mean Square Error (MMSE) beamformer, which is an array processing technique for enhancing ultrasound coherent plane wave synthesis (CPWC) image quality, typically approximates the minimum mean square error criterion MMSE using a Minimum Variance Distortionless Response (MVDR) beamformer and a scalar filter. In actual practice, scalar filters are typically approximated using a coherence factor whose calculation is based on the incident signal, which is estimated with errors. On the other hand, a minimum variance distortion free response (MVDR) beamformer generally obtains a data covariance matrix in the MVDR algorithm by a spatial smooth approximation estimation, which makes the calculated weight vector smaller, and thus the quality of the synthesized image is lower.
Disclosure of Invention
The purpose of the invention is as follows: to overcome the defects of the prior art, the invention provides an MMSE beam former, an MMSE beam forming method and a computer readable storage medium.
The technical scheme is as follows: the technical scheme provided by the invention is as follows:
an MMSE beam former is applied to coherent plane wave synthesis, receives echo data of plane waves of M different angles transmitted by a transducer through an N-line linear array, and performs beam synthesis; the echo data matrix is recorded as: u (n) ═ u1(n),u2(n),…,uM(n)]TThen the synthesized beam is:
wherein u isi(n) echo data representing a plane wave i,xi,j(n) echo data representing a plane wave i received at time instant n, jth array element; w is aDCR-MMSEAs a weight vector, wDCR-MMSEThe expression of (a) is:
wherein r isk(n) represents an intermediate parameter which is,qk(N) a set of snapshots of the N array element output signals of the beamformer, qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],For the diagonal loading parameters, I denotes the identity matrix and a denotes the time delay vector of all echo data.
An MMSE beamforming method comprising the steps of:
(1) exciting N element linear array transducers to emit M plane waves at different angles, and receiving echo data through N array elements;
(2) the output signal of each array element is acquired through snapshot and is recorded as qk(n),qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],Wherein x isi,j(n) echo data representing a plane wave i received at time instant n, jth array element;
(3) selecting diagonal line loading parameters according to requirements, and obtaining q according to the step (2)k(n) calculating a covariance matrix:
wherein I represents an identity matrix;
(4) defining an intermediate parameter rk(n),a represents the time delay of all echo dataVector quantity; q obtained according to step (2)k(n) and obtained in step (3)Calculate rk(n),
(5) Calculating a coherence factor:
(6) calculating a weight vector:
(7) the synthetic beam of the echo is calculated as:
a computer readable storage medium having stored thereon at least one instruction executable by a processor, the at least one instruction, when executed by the processor, implementing the MMSE beamforming method.
Has the advantages that: compared with the prior art, the invention has the following advantages: the MMSE beam forming scheme provided by the invention improves the technical scheme of the MMSE based on the Minimum Variance Distortionless Response (MVDR) beam former and a scalar filter to realize the approach of a synthesized beam to a minimum mean square error criterion, overcomes the technical problem of low quality of a synthesized image caused by covariance estimation and coherence factor calculation deviation in the existing Minimum Variance Distortionless Response (MVDR) beam forming method, can adaptively inhibit the influence of data noise and interference on imaging contrast and improve the quality of the synthesized image.
Drawings
Fig. 1 is a flowchart of an MMSE beamforming method according to the present invention.
FIG. 2 is a simulated image of 9 ideal lesions generated with different beamformers as involved in the example; wherein, fig. 2(a) is a simulated image of 9 ideal lesions generated by using a CPWC beamformer, with the value of M being 75; FIG. 2(b) is a simulated image of 9 ideal lesions generated using an SS-MVDR beamformer with M taken at 75; fig. 2(c) is a simulated image of 9 ideal lesions generated using a CPWC beamformer, with M taking the value 15; FIG. 2(d) is a simulated image of 9 ideal lesions generated using an SC-MVDR + CF beamformer with M taking the value of 15; FIG. 2(e) is a simulated image of 9 ideal lesions generated using an SS-MVDR + HRCF beamformer with M taking the value of 15; FIG. 2(f) is a simulated image of 9 ideal lesions generated by the joint SC-MVDR beamformer with M taking the value of 15; FIG. 2(g) is a simulated image of 9 ideal lesions generated by the DCR-MVDR beamformer with M being 15; FIG. 2(h) is a simulated image of 9 ideal lesions generated by the DCR-MMSE beamformer, where M is 15; fig. 2(a) to (h) all undergo log compression, showing a dynamic range of 60 dB.
FIG. 3 is an experimental image generated with different beamformers for resolution evaluation as referred to in the examples; fig. 3(a) is an experimental image generated by the CPWC beamformer, with the value of M being 75; FIG. 3(b) is an experimental image generated by the SS-MVDR beamformer with M taking the value of 75; fig. 3(c) is an experimental image generated by the CPWC beamformer, with the value of M being 15; FIG. 3(d) is an experimental image generated by the SC-MVDR + CF beamformer with a value of M of 15; FIG. 3(e) is an experimental image generated by the SS-MVDR + HRCF beamformer with M taking the value of 15; FIG. 3(f) is an experimental image generated by the joint SC-MVDR beamformer with a value of M of 15; FIG. 3(g) is an experimental image generated by the DCR-MVDR beamformer with the value of M being 15; FIG. 3(h) is an experimental image generated by the DCR-MMSE beamformer with the value of M being 15; fig. 3(a) to (h) are log compressed and displayed with a dynamic range of 70 dB.
Fig. 4 is a graph comparing the response of beams generated by a pair of CPWC beamformer (M75), SS-MVDR beamformer (M15) and CPWC beamformer (M15) on a scan centerline with nylon monofilament wire pairs having ordinates of 10mm, 30mm and 50mm, respectively; wherein, fig. 4(a) is a beam response contrast diagram on the cross section of the nylon monofilament metal wire at the ordinate 10mm on the scanning center line; FIG. 4(b) is a plot of the beam response versus a cross-section of a nylon monofilament wire at 30mm ordinate on the scan center line; fig. 4(c) is a plot of the beam response versus the cross-section of a nylon monofilament wire at 50mm ordinate on the scan center line.
Fig. 5 is a response comparison graph of beams generated by a nylon monofilament wire pair CPWC beamformer (M15), a joint SC-MVDR beamformer (M15), an SS-MVDR + CF beamformer (M15), an SS-MVDR + HRCF beamformer (M15), a DCR-MVDR beamformer (M15), and a DCR-MMSE beamformer (M15) at the scanning center line with ordinate of 10mm, 30mm, and 50mm, respectively; wherein, fig. 5(a) shows a response distribution comparison graph of a beam generated by a CPWC beamformer (M15) and a beam generated by a joint SC-MVDR beamformer (M15) on a cross section of a nylon monofilament metal wire at an ordinate of 10mm on a scanning center line; fig. 5(b) shows a graph of the response profile of the beam generated by the SS-MVDR + CF beamformer (M15) and the beam generated by the SS-MVDR + HRCF beamformer (M15) in a cross-section of a nylon monofilament wire at an ordinate of 30mm on the scan center line; fig. 5(c) shows a graph of response distributions of beams generated by the DCR-MVDR beamformer (M15) and beams generated by the DCR-MMSE beamformer (M15) in a cross-section of a nylon monofilament wire at an ordinate of 50mm on a scan center line.
Fig. 6 is an experimental image generated with different beamformers for evaluating contrast, where fig. 6(a) is an experimental image generated with a CPWC beamformer with M taking the value 75; FIG. 6(b) is an experimental image generated using an SS-MVDR beamformer with M taking the value 75; fig. 6(c) is an experimental image generated by using a CPWC beamformer, with the value of M being 15; FIG. 6(d) is an experimental image generated using an SC-MVDR + CF beamformer with a value of M of 15; FIG. 6(e) is an experimental image generated using a SS-MVDR + HRCF beamformer with the value of M being 15; FIG. 6(f) is an experimental image generated by the joint SC-MVDR beamformer with a value of M of 15; FIG. 6(g) is an experimental image generated by the DCR-MVDR beamformer with the value of M being 15; FIG. 6(h) is an experimental image generated by the DCR-MMSE beamformer with the value of M being 15; fig. 6(a) to (h) all undergo log compression, showing a dynamic range of 70 dB.
Fig. 7 is a carotid ultrasound image obtained by scanning the carotid artery with different beamformers, wherein fig. 7(a) is a carotid ultrasound image generated by using a CPWC beamformer with the M value being 75; FIG. 7(b) is a carotid ultrasound image generated using an SS-MVDR beamformer with M taken at 75; FIG. 7(c) is a carotid ultrasound image generated using a CPWC beamformer with M taking the value 15; FIG. 7(d) is a carotid ultrasound image generated using an SC-MVDR + CF beamformer with M taken as 15; FIG. 7(e) is a carotid ultrasound image generated using a SS-MVDR + HRCF beamformer with M taken at 15; FIG. 7(f) is a carotid ultrasound image generated by the combined SC-MVDR beamformer with M taken at 15; FIG. 7(g) is a carotid ultrasound image generated by the DCR-MVDR beamformer with M taken as 15; FIG. 7(h) is a carotid ultrasound image generated by the DCR-MMSE beamformer with M taken as 15; fig. 7(a) to (h) all undergo log compression, showing a dynamic range of 60 dB.
FIG. 8 is a table of experimental results referred to in the examples.
Detailed Description
The invention will be further described with reference to the accompanying drawings and specific embodiments. It is to be understood that the present invention may be embodied in various forms, and that there is no intention to limit the invention to the specific embodiments illustrated, but on the contrary, the intention is to cover some exemplary and non-limiting embodiments shown in the attached drawings and described below.
It is to be understood that the features listed above for the different embodiments may be combined with each other to form further embodiments within the scope of the invention, where technically feasible. Furthermore, the particular examples and embodiments of the invention described are non-limiting, and various modifications may be made in the structure, steps, and sequence set forth above without departing from the scope of the invention.
The invention provides an MMSE (minimum mean square error) beam former which is applied to coherent plane wave synthesis, wherein the beam former receives echo data of plane waves of M different angles transmitted by a transducer through an N-line linear array and performs beam synthesis; the echo data matrix is recorded as: u (n) ═ u1(n),u2(n),…,uM(n)]TThen, then are combinedThe beam forming is:
wherein u isi(n) echo data representing a plane wave i,xi,j(n) echo data representing a plane wave i received at time instant n, jth array element; w is aDCR-MMSEAs a weight vector, wDCR-MMSEThe expression of (a) is:
wherein r isk(n) represents an intermediate parameter which is,qk(N) a set of snapshots of the N array element output signals of the beamformer, qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],For the diagonal loading parameters, I denotes the identity matrix and a denotes the time delay vector of all echo data.
The present invention also provides an MMSE beam forming method, the flow of which is shown in fig. 1, including the steps of:
(1) exciting N element linear array transducers to emit M plane waves at different angles, and receiving echo data through N array elements;
(2) by capturing each element by snap-shotOutput signal, noted qk(n),qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],Wherein x isi,j(n) echo data representing a plane wave i received at time instant n, jth array element;
(3) selecting diagonal line loading parameters according to requirements, and obtaining q according to the step (2)k(n) calculating a covariance matrix:
wherein I represents an identity matrix;
(4) defining an intermediate parameter rk(n),a represents the time delay vector of all echo data; q obtained according to step (2)k(n) and obtained in step (3)Calculate rk(n),
(5) Calculating a coherence factor:
(6) calculating a weight vector:
(7) the synthetic beam of the echo is calculated as:
the following illustrates the design principles of the present invention.
A. Coherent plane composite CPWC
In the CPWC, the echo data of the plane waves emitted at different angles are acquired by emitting the plane waves at a plurality of angles and beam-forming is performed to obtain an ultrasound image. Assuming that a linear array comprises N array elements, and the linear array transmits M plane waves with different angles, echo data received by the linear array is represented by a two-dimensional matrix:
wherein x isi,j(n) echo data representing a plane wave i received at time instant n, jth array element;
the coherent composite beam for obtaining the echo data is:
B. MVDR beam former with space smoothing function
The MVDR algorithm can be integrated into the coherent combining process, and not only can combine the data in each plane wave transmission (MVDR upon reception), but also can combine the data at different emission angles (MVDR upon emission).
An MVDR beamformer is used, echo data of each plane wave emitted by the DAS beamformer forms a corresponding low-resolution image, and then the low-resolution images are combined through an MVDR algorithm to obtain a composite image. The echo data of each plane wave is recorded as uiThen the echo data matrix is represented as: u (n) ═ u1(n),u2(n),…,uM(n)]TWherein u isi(n) echo data of plane wave i, uiThe expression of (n) is:
during source propagation we observe the presence of spatial correlation at different reception points (sensor elements). Whereas in ultrasound imaging the light source is not self-radiating, but is formed by reflection of the incident light beam. A general MVDR beamformer processes echo signals of a single shot plane wave. However, for the current array, the components generated by the reflection of different plane waves included in the input vector u (N) can be described by the reciprocity of acoustics theorem, and based on this, u (N) can be regarded as the echo data of M plane waves emitted from different angles received by N linear arrays.
To find a beamformer to be applied to u (n), we modeled u (n):
u(n)=spa+v(n) (4)
wherein s ispRepresenting signal amplitude, v (n) is a combination of noise and non-target interference, and a represents a time delay vector.
Then, a beam forming weight vector calculation model is given:
the model satisfies the constraint: w is aHa=1;
Solving the formula (5) to obtain a weight vector:
wherein R isvIs a noise plus interference covariance matrix;
in practice, we usually use a data covariance matrix R containing the desired signaluReplacement of RvObtaining:
the output signal of the beamformer is then:
in general, we passObtaining a data covariance matrix R in the MVDR algorithm by space smooth approximate estimationuHowever, this approximation method results in a calculated weight vector wMVDRTo eliminate this effect, w needs to be reducedMVDRBy multiplying by the average of the elements in u (n), we turn this beamformer into a spatially smooth beamformer.
C. Coherence factor
In the prior art, the MVDR and the coherence factor are generally used in combination to enhance the performance of composite imaging. The coherence factor CF is defined as representing the ratio of the sum of the amount of coherent data to the sum of the amount of incoherent data in the data vector. This way of combining MVDR with coherence factors can be used to approximate the minimum mean square error criterion MMSE. The formula for calculating the coherence factor CF is:
wherein the content of the first and second substances,the value of CF ranges from 0 to 1, the lower the value of CF as the coherence of the signal received at the transducer element is lower. Based on this property, we use CF as a post-filter to suppress blurring artifacts. To form a better filtered scalar, CF is modified to create two variants: high-resolution coherence coefficient (HRCF) and scale coefficient (scCF):
using y in the formulaMVDR(n) replaces yavg(n), to understand the effect of doing so on the imaging results, we note yavg(n) weight vectorAlso satisfies the corresponding condition of no distortionIn such a beamforming weight vector, wMVDRSatisfy the minimum output power or yMVDR(n)|2≤|yavg(n)|2. Thus, the high resolution coherence factor can reduce the effect of the post-filter scalar on the imaging contrast.
The scaling factor (scCF) is expressed as:
η is a parameter that ranges from 0 to 1, scCF (n) is less than 1 but greater than CF. therefore, it reduces the effect of the post-filter scalar on the imaging contrast and makes the image synthesis result closer to that of the MVDR beamformer.
D. MVDR algorithm based on space consistency
In the invention, a new method is developed to estimate the covariance of u (n), and a plurality of snapshots q are collectedk(n) covariance for estimating u (n), qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],u (n) and qk(n) statistical similarity between them complies with the Van Satt-Zernike theorem. We first detect the spatial coherence between the elements in u (N), and based on the acoustic reciprocity of coherent plane wave complex imaging (CPWC) when receiving signals, we consider u (N) as the backscattered signals from point targets received by N linear arrays, and then use the fourier transform of the beam pattern and the distance between two elements in the array to calculate the spatial coherence between the corresponding two signals in u (N) by using the van-zernike theorem. Compared with u (n), qk(N) is the back scattered signal from the point target received by all but the k-th element of the N element linear arrays, when N is large enough, we consider u (N) and qkThe difference between (n) is small enough for u (n) and qk(n) have similar second order statistics, therefore, we estimate the covariance matrix using the following formulaArraying:
wherein, I represents an identity matrix for diagonal loading parameters.
e MMSE beamformer
In the invention, a spatial coherence method is expanded, and a new MMSE beam former suitable for CPWC is developed. We first assume that the main signal is deterministic and that the backscatter data is a narrowband signal. The output signals y and s of the beamformer are according to equation (4)pThe mean square error MSE between is calculated as:
E{|y-sp|2)=wHRvw+|sp|2|1-wHa|2(15)
where E represents the expected value operator.
According to the minimum mean square error MSE, calculating a weight vector as follows:
further obtaining:
of these, β(s)p) As a post-filter scalar, β(s)p) The expression of (a) is:
substituting formula (17) for formula (15) yields:
for w in formula (6)MVDRWe pass the MSEMVDRError between the main signal and the MVDR beamformer output signal is described:
comparing equation (19) with equation (20), we find MSEout≤MSEMVDRThus for the main signal spThe output of the MMSE beamformer is more accurate.
However, according to equation (18), calculating the weight vector of MMSE also requires estimating spAndthe output signal of the MVDR beamformer is spIs estimated unbiased, andand MSEMVDRDirect correlation (e.g., equation (20)), and thus β(s)p) Can be expressed in terms of the mean square error of the output signal of the MVDR beamformer. Based on this analysis, we used the DCR-MVDR beamformer and the snapshot set q in equation (12)k(n) to estimate β(s)p)。
In the space phase drying method, u (n) and qk(n) have similar covariance matrices, and therefore,can be regarded as qk(n) covariance matrix, and wDCR-MVDRCan be used as weight vector of MVDR beam former. ByIn thatAndmean value s ofpThe same applies, therefore,andso β(s)p) May be used with qk(n) is estimated as the mean square error of the input output signal of the DCR-MVDR beamformer.
Order toBy { rkThe mean and variance of (n) are substituted into (18) as spAndto obtain an approximation of:
in the form of a coherence factor, butAre calculated from the output data of the beamformer rather than from the input data. The new MMSE beamformer has weight vectors ofThe output signal is:
the above is the design principle of the MMSE beamformer and beamforming method proposed by the present invention.
Further, the present invention also provides a computer-readable storage medium storing at least one instruction executable by a processor, where the at least one instruction, when executed by the processor, implements the MMSE beamforming method.
The technical effects of the present invention are illustrated by specific experimental data below.
We first generated simulated images of 9 ideal lesions through the beamformer to assess the resolution of the images generated by the different beamformers. FIG. 2 shows simulated images of 9 ideal lesions generated with different beamformers; wherein, fig. 2(a) is a simulated image of 9 ideal lesions generated by using a CPWC beamformer, with the value of M being 75; FIG. 2(b) is a simulated image of 9 ideal lesions generated using an SS-MVDR beamformer with M taken at 75; fig. 2(c) is a simulated image of 9 ideal lesions generated using a CPWC beamformer, with M taking the value 15; FIG. 2(d) is a simulated image of 9 ideal lesions generated using an SC-MVDR + CF beamformer with M taking the value of 15; FIG. 2(e) is a simulated image of 9 ideal lesions generated using an SS-MVDR + HRCF beamformer with M taking the value of 15; FIG. 2(f) is a simulated image of 9 ideal lesions generated by the joint SC-MVDR beamformer with M taking the value of 15; FIG. 2(g) is a simulated image of 9 ideal lesions generated by the DCR-MVDR beamformer with M being 15; FIG. 2(h) is a simulated image of 9 ideal lesions generated by the DCR-MMSE beamformer, where M is 15; fig. 2(a) to (h) all undergo log compression, showing a dynamic range of 60 dB.
In fig. 2(a) to (h), the lesions are uniformly distributed, with centers located at 17mm, 30mm and 43mm in the longitudinal direction and at-12 mm, Omm and 12mm in the transverse direction. Fig. 2(a) and 2(b) are images generated based on a full dataset (the full dataset is plane wave echo data at 75 different angles), whereas fig. (c) to (h) employ only plane wave echo data at 15 angles.
Comparing fig. 2(a) and fig. 2(b), it can be seen that the SS-MVDR beamformer is effective in improving the resolution of the lesion compared to the CPWC beamformer. Comparing fig. 2(c) to (h), when M is 15, the resolution of the CPWC beamformer is lowest (as in fig. 2(c)), and the DCR-MVDR beamformer has a higher resolution (as in fig. 2(e)), especially at a depth of 17mm at the transducer surface (near field region). The resolution of the SC-MVDR + CF beamformer, the SS-MVDR + HRCF beamformer, and the joint SC-MVDR beamformer is comparable to the resolution of the DCR-MVDR beamformer. However, by carefully observing these images, we find that some artifacts are introduced inside the lesion located in the near field region or at the edge of the image, while the DCR-MMSE beamformer proposed by the present invention has the highest image quality, the DCR-MMSE beamformer can not only improve the resolution of the lesion, but also suppress the artifacts, further improving the image quality.
For fig. 2(a) to 2(h), we use the following formula to quantify the image quality:
wherein IinAnd IoutThe average intensity (in decibels) measured inside and outside the focus respectively, the value range of CR is 0 to 1, and the larger the value of CR, the higher the resolution is. The table shown in fig. 8 shows the image quality average values in fig. 2(a) to 2 (h). From the calculation results, the DCR-MMSE beamformer has the highest peak response compared to the other beamformers.
Next, we scan the tissue phantom with different beamformers, yielding experimental data for evaluating the spatial resolution performance of the beamformers. Setting the sound velocity of the scanning to 1540 +/-10 m/s and the slope of the background attenuation coefficient to 0.5dBcm- 1MH-1. The scanned objects are 7 nylon monofilament metal wires with the diameter of 100 mm which are hung in the mottled background, the 7 nylon monofilament metal wires are arranged in parallel, and the beam emission direction is parallel to the length direction of the nylon monofilament metal wires during scanning. One end of the nylon monofilament metal wire is regarded as a point, and after scanning, 5 points are positioned in the scanningOn the axis (i.e. the abscissa on the scanned image is 0), the 5 points have the ordinate positions of 10mm, 20 mm, 30mm, 40 mm and 50mm on the scanned image, and the other two points are located on the abscissa of-10 mm and the ordinate of 40 mm on the scanned image, and the other point is located on the abscissa of 10mm and the ordinate of 40 mm on the scanned image. Specific scanning data is shown in fig. 3, fig. 3(a) is an experimental image generated by a CPWC beamformer, and M takes a value of 75; FIG. 3(b) is an experimental image generated by the SS-MVDR beamformer with M taking the value of 75; fig. 3(c) is an experimental image generated by the CPWC beamformer, with the value of M being 15; FIG. 3(d) is an experimental image generated by the SC-MVDR + CF beamformer with a value of M of 15; FIG. 3(e) is an experimental image generated by the SS-MVDR + HRCF beamformer with M taking the value of 15; FIG. 3(f) is an experimental image generated by the joint SC-MVDR beamformer with a value of M of 15; FIG. 3(g) is an experimental image generated by the DCR-MVDR beamformer with the value of M being 15; FIG. 3(h) is an experimental image generated by the DCR-MMSE beamformer with the value of M being 15; fig. 3(a) to (h) are log compressed and displayed with a dynamic range of 70 dB.
Fig. 3(a) and 3(b) are experimental images obtained from full-angle data, and from fig. 3, the SS-MVDR beamformer has a higher line resolution than the CPWC beamformer. Fig. 3(c) to 3(h) are beamformers calculated based on echo data of 15 angles, wherein the line resolution of the remaining beamformers, except the SC-MVDR + CF beamformer, is higher than that of the CPWC beamformer (fig. 3 (c)). Due to the coherence of the signals received by the individual beamformer array elements, coherent spots may be formed in the imaging region, resulting in artifacts in the entire imaging region, thereby affecting image quality.
The DCR-MMSE (in fig. 3 (h)) achieves the highest line resolution, but it is equivalent to the line resolution obtained with DCR-MVDR (in fig. 3 (g)). Whereas the line resolution of MVDR + CF and MVDR + HRCF is better than that of CPWC, but worse than that of DCR-MMSE.
We present the beam profiles of the above-described beamformer in fig. 4 and 5. Fig. 4 is a graph comparing the response of beams generated by a pair of CPWC beamformer (M75), SS-MVDR beamformer (M15) and CPWC beamformer (M15) on a scan centerline with nylon monofilament wire pairs having ordinates of 10mm, 30mm and 50mm, respectively; wherein, fig. 4(a) is a beam response contrast diagram on the cross section of the nylon monofilament metal wire at the ordinate 10mm on the scanning center line; FIG. 4(b) is a plot of the beam response versus a cross-section of a nylon monofilament wire at 30mm ordinate on the scan center line; fig. 4(c) is a plot of the beam response versus the cross-section of a nylon monofilament wire at 50mm ordinate on the scan center line. As can be seen from fig. 4, the reduction of the transmitted plane wave (compare M75 and M15) results in an increase of the side lobes of the response signal, which is particularly significant in the near field region.
Fig. 5 is a response comparison graph of beams generated by a nylon monofilament wire pair CPWC beamformer (M15), a joint SC-MVDR beamformer (M15), an SS-MVDR + CF beamformer (M15), an SS-MVDR + HRCF beamformer (M15), a DCR-MVDR beamformer (M15), and a DCR-MMSE beamformer (M15) at the scanning center line with ordinate of 10mm, 30mm, and 50mm, respectively; wherein, fig. 5(a) shows a response distribution comparison graph of a beam generated by a CPWC beamformer (M15) and a beam generated by a joint SC-MVDR beamformer (M15) on a cross section of a nylon monofilament metal wire at an ordinate of 10mm on a scanning center line; fig. 5(b) shows a graph of the response profile of the beam generated by the SS-MVDR + CF beamformer (M15) and the beam generated by the SS-MVDR + HRCF beamformer (M15) in a cross-section of a nylon monofilament wire at an ordinate of 30mm on the scan center line; fig. 5(c) shows a graph of response distributions of beams generated by the DCR-MVDR beamformer (M15) and beams generated by the DCR-MMSE beamformer (M15) in a cross-section of a nylon monofilament wire at an ordinate of 50mm on a scan center line.
As can be seen from fig. 5, when the number of transmitted plane waves is all 15, the main lobe of the DCR-MMSE beamformer is narrowest, indicating that its radiation is most concentrated. We measure the pulse width of all responses with FWHM, calculate the mean FWHM of each beamformer, and record the calculation results in the table shown in fig. 8.
To evaluate contrast resolution, we scanned two anechoic cysts in the speckle background with the beamformer described above, and the obtained scan data is shown in fig. 6, where fig. 6(a) is an experimental image generated with a CPWC beamformer and M is 75; FIG. 6(b) is an experimental image generated using an SS-MVDR beamformer with M taking the value 75; fig. 6(c) is an experimental image generated by using a CPWC beamformer, with the value of M being 15; FIG. 6(d) is an experimental image generated using an SC-MVDR + CF beamformer with a value of M of 15; FIG. 6(e) is an experimental image generated using a SS-MVDR + HRCF beamformer with the value of M being 15; FIG. 6(f) is an experimental image generated by the joint SC-MVDR beamformer with a value of M of 15; FIG. 6(g) is an experimental image generated by the DCR-MVDR beamformer with the value of M being 15; FIG. 6(h) is an experimental image generated by the DCR-MMSE beamformer with the value of M being 15; fig. 6(a) to (h) all undergo log compression, showing a dynamic range of 70 dB.
As can be seen from fig. 6, the diameter of the cyst is about 3mm, and the two cysts are located at positions 15 mm and 45 mm on the ordinate of the scanning center line, respectively. As can be seen from fig. 6, the DCR-MMSE beamformer generates experimental images with optimal contrast for the two cysts. We calculate the mean value of the image quality CR of each cyst, and then record the calculation result in the table shown in fig. 8.
Fig. 7 is a carotid ultrasound image obtained by scanning the carotid artery with different beamformers, wherein fig. 7(a) is a carotid ultrasound image generated by using a CPWC beamformer with the M value being 75; FIG. 7(b) is a carotid ultrasound image generated using an SS-MVDR beamformer with M taken at 75; FIG. 7(c) is a carotid ultrasound image generated using a CPWC beamformer with M taking the value 15; FIG. 7(d) is a carotid ultrasound image generated using an SC-MVDR + CF beamformer with M taken as 15; FIG. 7(e) is a carotid ultrasound image generated using a SS-MVDR + HRCF beamformer with M taken at 15; FIG. 7(f) is a carotid ultrasound image generated by the combined SC-MVDR beamformer with M taken at 15; FIG. 7(g) is a carotid ultrasound image generated by the DCR-MVDR beamformer with M taken as 15; FIG. 7(h) is a carotid ultrasound image generated by the DCR-MMSE beamformer with M taken as 15; fig. 7(a) to (h) all undergo log compression, showing a dynamic range of 60 dB.
Under normal circumstances, the artery appears as a circular structure on the sonogram, with the lumen appearing as an anechoic region surrounded by specular scattering from the artery wall. The intra-cavity contrast was calculated for each image and the calculation was recorded in the table of fig. 8.
In fig. 7, the carotid ultrasound image generated by the CPWC beamformer (fig. 7(a)) and the carotid ultrasound image generated by the SS-MVDR beamformer (fig. 7(b)) are both generated with complete data sets, and the image quality is relatively close. The image quality of carotid ultrasound images generated by the DCR-MVDR beamformer is significantly higher than that of the CPWC beamformer and the SS-MVDR beamformer. When M is 15, the contrast of the carotid ultrasound image generated by the CPWC beamformer and the contrast of the carotid ultrasound image generated by the DCR-MVDR beamformer are both low. Between the SC-MVDR + CF beam former and the joint SC-MVDR beam former, the contrast of the carotid artery ultrasonic image generated by the joint SC-MVDR beam former is higher, even exceeds the DCR-MMSE beam former, but the carotid artery ultrasonic image generated by the joint SC-MVDR beam former has an artifact in a near field area, so that the image quality is poor. The SC-MVDR + CF beamformer and the joint SC-MVDR beamformer can also improve the imaging contrast to some extent, but there are also artifacts in the near field region. While the DCR-MMSE beamformer can provide both high imaging contrast and high image quality, as can be seen in fig. 7(h), the entire lumen and artery boundaries are very clear.
The above description is only of the preferred embodiments of the present invention, and it should be noted that: it will be apparent to those skilled in the art that various modifications and adaptations can be made without departing from the principles of the invention and these are intended to be within the scope of the invention.
Claims (3)
1. An MMSE beam former is applied to coherent plane wave synthesis and is characterized in that the beam former receives echo data of plane waves of M different angles transmitted by a transducer through an N-line linear array and carries out beam synthesis; the echo data matrix is recorded as: u (n) ═ u1(n),u2(n),…,uM(n)]TThen the synthesized beam is:
wherein u isi(n) echo data representing a plane wave i,xi,j(n) echo data representing a plane wave i received at time instant n, jth array element; w is aDCR-MMSEAs a weight vector, wDCR-MMSEThe expression of (a) is:
2. An MMSE beamforming method, comprising the steps of:
(1) exciting N element linear array transducers to emit M plane waves at different angles, and receiving echo data through N array elements;
(2) the output signal of each array element is acquired through snapshot and is recorded as qk(n),qk(n)=[qk,1(n),qk,2(n),…,qk,N(n)],Wherein x isi,j(n) echo data representing a plane wave i received at time instant n, jth array element;
(3) selecting diagonal line loading parameters according to requirements, and obtaining q according to the step (2)k(n) calculating a covariance matrix:
wherein I represents an identity matrix;
(4) defining an intermediate parameter rk(n),a represents the time delay vector of all echo data; q obtained according to step (2)k(n) and obtained in step (3)Calculate rk(n),
(5) Calculating a coherence factor:
(6) calculating a weight vector:
(7) the synthetic beam of the echo is calculated as:
3. a computer-readable storage medium having stored thereon at least one instruction executable by a processor, the at least one instruction when executed by the processor implementing the MMSE beamforming method of claim 2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010288748.1A CN111466949B (en) | 2020-04-13 | 2020-04-13 | MMSE beam forming device, MMSE beam forming method and computer readable storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010288748.1A CN111466949B (en) | 2020-04-13 | 2020-04-13 | MMSE beam forming device, MMSE beam forming method and computer readable storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111466949A true CN111466949A (en) | 2020-07-31 |
CN111466949B CN111466949B (en) | 2023-06-27 |
Family
ID=71752403
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010288748.1A Active CN111466949B (en) | 2020-04-13 | 2020-04-13 | MMSE beam forming device, MMSE beam forming method and computer readable storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111466949B (en) |
Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050228279A1 (en) * | 2004-03-31 | 2005-10-13 | Siemens Medical Solutions Usa, Inc. | Coherence factor adaptive ultrasound imaging methods and systems |
CN102835975A (en) * | 2012-09-19 | 2012-12-26 | 重庆博恩克医疗设备有限公司 | MV (Minimum variance) wave beam formation and MV-based CF (correlation factor) fusion method |
CN103245956A (en) * | 2013-04-19 | 2013-08-14 | 哈尔滨工程大学 | GPS (global positioning system) multipath mitigation method based on robust beam forming algorithm |
CN103536316A (en) * | 2013-09-22 | 2014-01-29 | 华中科技大学 | Method for self-adaptation ultrasonic imaging of spatio-temporally smoothed coherence factor type |
CN103941234A (en) * | 2014-04-24 | 2014-07-23 | 清华大学 | Beam forming device design method for multipath signals |
CN106510761A (en) * | 2016-12-12 | 2017-03-22 | 重庆大学 | Signal-noise-ratio-post-filtering-and-characteristic-space-fusion minimum-variance ultrasonic imaging method |
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 |
CN108354626A (en) * | 2018-03-31 | 2018-08-03 | 华南理工大学 | A variety of MV high clearing systems fast medical ultrasonic image systems based on GPU |
CN108403148A (en) * | 2018-04-16 | 2018-08-17 | 武汉维视医学影像有限公司 | A kind of ultrasonic CT imaging process based on MV Adaptive beamformers |
CN108836389A (en) * | 2018-06-01 | 2018-11-20 | 西安交通大学 | Plane wave reference point is concerned with adaptive beam compound imaging method |
CN109164453A (en) * | 2018-10-25 | 2019-01-08 | 国网内蒙古东部电力有限公司检修分公司 | A kind of minimum variance ultrasonic imaging method merging highly coherent filter |
CN109245814A (en) * | 2018-09-13 | 2019-01-18 | 哈尔滨工业大学 | Adaptive beamformer method based on maximum likelihood resampling |
US20190079825A1 (en) * | 2017-09-08 | 2019-03-14 | The Aerospace Corporation | INTERFERENCE SUPPRESSION USING REPEATED REDUCED RANK ADAPTIVE FILTERING IN FRACTIONAL FOURIER TRANSFORM (FrFT) DOMAINS |
CN109754407A (en) * | 2019-01-10 | 2019-05-14 | 青岛海信医疗设备股份有限公司 | Ultrasound image processing method, device and equipment |
CN110554393A (en) * | 2019-07-12 | 2019-12-10 | 华南理工大学 | high-contrast minimum variance imaging method based on deep learning |
-
2020
- 2020-04-13 CN CN202010288748.1A patent/CN111466949B/en active Active
Patent Citations (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050228279A1 (en) * | 2004-03-31 | 2005-10-13 | Siemens Medical Solutions Usa, Inc. | Coherence factor adaptive ultrasound imaging methods and systems |
CN102835975A (en) * | 2012-09-19 | 2012-12-26 | 重庆博恩克医疗设备有限公司 | MV (Minimum variance) wave beam formation and MV-based CF (correlation factor) fusion method |
CN103245956A (en) * | 2013-04-19 | 2013-08-14 | 哈尔滨工程大学 | GPS (global positioning system) multipath mitigation method based on robust beam forming algorithm |
CN103536316A (en) * | 2013-09-22 | 2014-01-29 | 华中科技大学 | Method for self-adaptation ultrasonic imaging of spatio-temporally smoothed coherence factor type |
CN103941234A (en) * | 2014-04-24 | 2014-07-23 | 清华大学 | Beam forming device design method for multipath signals |
CN106510761A (en) * | 2016-12-12 | 2017-03-22 | 重庆大学 | Signal-noise-ratio-post-filtering-and-characteristic-space-fusion minimum-variance ultrasonic imaging method |
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 |
US20190079825A1 (en) * | 2017-09-08 | 2019-03-14 | The Aerospace Corporation | INTERFERENCE SUPPRESSION USING REPEATED REDUCED RANK ADAPTIVE FILTERING IN FRACTIONAL FOURIER TRANSFORM (FrFT) DOMAINS |
CN108354626A (en) * | 2018-03-31 | 2018-08-03 | 华南理工大学 | A variety of MV high clearing systems fast medical ultrasonic image systems based on GPU |
CN108403148A (en) * | 2018-04-16 | 2018-08-17 | 武汉维视医学影像有限公司 | A kind of ultrasonic CT imaging process based on MV Adaptive beamformers |
CN108836389A (en) * | 2018-06-01 | 2018-11-20 | 西安交通大学 | Plane wave reference point is concerned with adaptive beam compound imaging method |
CN109245814A (en) * | 2018-09-13 | 2019-01-18 | 哈尔滨工业大学 | Adaptive beamformer method based on maximum likelihood resampling |
CN109164453A (en) * | 2018-10-25 | 2019-01-08 | 国网内蒙古东部电力有限公司检修分公司 | A kind of minimum variance ultrasonic imaging method merging highly coherent filter |
CN109754407A (en) * | 2019-01-10 | 2019-05-14 | 青岛海信医疗设备股份有限公司 | Ultrasound image processing method, device and equipment |
CN110554393A (en) * | 2019-07-12 | 2019-12-10 | 华南理工大学 | high-contrast minimum variance imaging method based on deep learning |
Non-Patent Citations (5)
Title |
---|
NGHIA NGUYEN, RICHARD PRAGER: "Minimum variance beamformers for", 《PROC. SPIE 10139, MEDICAL IMAGING》 * |
NGHIA Q. NGUYEN AND RICHARD W. PRAGER: "A Spatial Coherence Approach to Minimum Variance Beamforming for Plane-Wave Compounding", 《IEEE TRANSACTIONS ON ULTRASONICS,FERROELECTRICS,AND FREQUENCY CONTROL》 * |
NGHIA Q. NGUYEN AND RICHARD W. PRAGER: "Mean-Squared Error Beamforming for Coherent Plane-Wave Compounding", 《2019 IEEE INTERNATIONAL ULTRASONICS SYMPOSIUM (IUS)》 * |
NGHIA Q. NGUYEN AND RICHARD W. PRAGER: "Minimum Variance Approaches to Ultrasound", 《IEEE TRANSACTIONS ON MEDICAL IMAGING》 * |
NGHIA Q. NGUYEN, RICHARD W. PRAGER AND MICHAEL F. INSANA: "Improvements to ultrasonic beamformer design and implementation derived from the", 《THE JOURNAL OF THE ACOUSTICAL SOCIETY OF AMERICA》 * |
Also Published As
Publication number | Publication date |
---|---|
CN111466949B (en) | 2023-06-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Xu et al. | Spatio-temporally smoothed coherence factor for ultrasound imaging [correspondence] | |
Kortbek et al. | Sequential beamforming for synthetic aperture imaging | |
Shin et al. | Spatial prediction filtering of acoustic clutter and random noise in medical ultrasound imaging | |
US20130343627A1 (en) | Suppression of reverberations and/or clutter in ultrasonic imaging systems | |
KR20210042907A (en) | Method and system for non-invasive characterization of heterogeneous media using ultrasound | |
EP2820445A2 (en) | Clutter suppression in ultrasonic imaging systems | |
Morgan et al. | Multi-covariate imaging of sub-resolution targets | |
Agarwal et al. | Improving spatial resolution using incoherent subtraction of receive beams having different apodizations | |
Tasnim et al. | Study of speckle noise reduction from ultrasound B-mode images using different filtering techniques | |
US11413020B2 (en) | Deep tissue super-resolution ultrasound imaging method and system | |
Touil et al. | Analysis of motion tracking in echocardiographic image sequences: Influence of system geometry and point-spread function | |
Thon et al. | Detection of point scatterers in medical ultrasound | |
CN111466949B (en) | MMSE beam forming device, MMSE beam forming method and computer readable storage medium | |
JP7237116B2 (en) | Method and system for estimating ultrasound attenuation parameters | |
Santos et al. | Evaluation of coherence-based beamforming for B-mode and speckle tracking echocardiography | |
KR102452220B1 (en) | Imaging method, apparatus implementing the method, a computer program and a computer-readable storage medium | |
US20220091243A1 (en) | Systems and Methods for Ultrasound Attenuation Coefficient Estimation | |
Klimonda et al. | Spatial and frequency compounding in application to attenuation estimation in tissue | |
US20230304850A1 (en) | Method and system for evaluating an assumed wave propagation speed in a medium | |
Guenther et al. | Broadband optimal contrast resolution beamforming | |
Opretzka et al. | A model-based synthetic aperture image reconstruction technique for high-frequency ultrasound | |
Wennerstrom et al. | An iterative synthetic aperture imaging algorithm with correction of diffraction effects | |
Lou et al. | A K-space-based approach to coherence estimation | |
Nguyen et al. | Mean-Squared Error Beamforming for Coherent Plane-Wave Compounding | |
Thanoon et al. | A New Discrimination Method of Imaging Regions for Improved Ultrafast Ultrasound Imaging Performance |
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 |