CN105310727B - Tissue elasticity imaging method and graphics processor - Google Patents
Tissue elasticity imaging method and graphics processor Download PDFInfo
- Publication number
- CN105310727B CN105310727B CN201510790828.6A CN201510790828A CN105310727B CN 105310727 B CN105310727 B CN 105310727B CN 201510790828 A CN201510790828 A CN 201510790828A CN 105310727 B CN105310727 B CN 105310727B
- Authority
- CN
- China
- Prior art keywords
- ultrasonic signal
- data window
- estimated value
- displacement
- cross
- 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.)
- Active
Links
Landscapes
- Ultra Sonic Daignosis Equipment (AREA)
- Image Analysis (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
A kind of tissue elasticity imaging method of present invention offer and graphics processor, this method include：Receive the first ultrasonic signal and the second ultrasonic signal before and after tissue generation deformation；The parameter that first ultrasonic signal, the second ultrasonic signal, cross correlation algorithm include is deposited into the corresponding global storages of GPU；It is that the corresponding processing thread of each data window distribution obtains the corresponding Displacement Estimation value of each data window by the cross-correlation coefficient of the first ultrasonic signal and the second ultrasonic signal in lineation journey parallel computation corresponding data window everywhere according to the number of data window；Each displacement estimated value is filtered respectively, obtains the corresponding strain estimated value of each displacement estimated value；Imaging is carried out to each strain estimated value, obtains the elastogram result of tissue.By way of multi-threading parallel process, while ensureing tissue elasticity imaging results accuracy, hence it is evident that improve the processing speed of elastogram.
Description
Technical field
The invention belongs to technical field of medical equipment, it is specifically related to a kind of tissue elasticity imaging method and graphics process
Device.
Background technology
The elasticity of tissue is to be influenced maximum biological tissue's mechanics parameter, a large amount of physiology of human body by physiology and pathological factor
Variation pathologically is all along with the variation of tissue elasticity, such as with the intensification of hepatic fibrosis-renal tubular ectasia syndrome degree, the hardness meeting of liver
It becomes larger.It therefore can be by the elasticity of tissue as an important parameter for reacting biological properties.The 1990s
First Japanese scholars Y.Yamakoshi and American scholar J.Ophir first proposed Ultrasonic Elasticity Imaging, and the technology is to organize
The elastic parameters such as modulus of shearing, Young's modulus, stress, strain be imaging object.
Ultrasonic elastograph imaging obtains the ultrasound echo signal that group is woven in before and after generating deformation first with ultrasonic scanning system,
And then the ultrasound echo signal is analyzed, the elastic parameter of tissue is obtained, and then imaging is carried out to the elastic parameter,
The elastic size of tissue is intuitively shown by the color of image.
During carrying out ultrasonic elastograph imaging processing to tissue, it is mainly concerned with the position that group is woven under certain excitation
It moves, strain estimation procedure, calculation amount is larger, wherein the cross correlation algorithm in Displacement Estimation consumes a large amount of time.Why
Displacement Estimation is carried out using cross correlation algorithm, is primarily due to relative to phase zero estimation, joint auto-correlation, frequency spectrum strain estimation
Scheduling algorithm, cross correlation algorithm have preferably estimation accuracy, can improve elastogram quality.Therefore how higher in guarantee
The processing speed that elastogram is improved under the premise of image quality, becomes a urgent problem to be solved.
Invention content
It is at least one in order to solve the problems, such as to mention in background technology, the present invention provide a kind of tissue elasticity imaging method and
Graphics processor, while ensureing elastogram quality, to improve the processing speed of elastogram.
The present invention provides a kind of tissue elasticity imaging methods, including：
Step 1：After receiving the first ultrasonic signal and the tissue generation deformation before the tissue generation deformation
The second ultrasonic signal；
Step 2：First ultrasonic signal, second ultrasonic signal and user is pre-set mutually
Related parameter is deposited into the corresponding global storages of graphics processor GPU, wherein the mutual related parameter is for carrying out displacement
The parameter for including in cross correlation algorithm used by estimation, the mutual related parameter includes the initial position of data window, data window
Size, search range；
Step 3：According to the number of the data window of the user setting, corresponding processing thread is distributed for each data window,
And pass through the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere and second ultrasonic signal
Cross-correlation coefficient obtains the corresponding Displacement Estimation value of each data window；
Step 4：Each displacement estimated value is filtered respectively, obtains the corresponding strain of each displacement estimated value
Estimated value；
Step 5：Imaging is carried out to each strain estimated value, obtains the elastogram result of the tissue.
The present invention provides a kind of graphics processors, including：
Receiving module is generated for receiving the tissue described in the first ultrasonic signal before deformation and tissue generation
The second ultrasonic signal after deformation；
Processing module is stored, for first ultrasonic signal, second ultrasonic signal and user is advance
The mutual related parameter being arranged is deposited into the corresponding global storages of graphics processor GPU, wherein the mutual related parameter is to use
The parameter for including in the cross correlation algorithm used by carrying out Displacement Estimation, the mutual related parameter include the start bit of data window
Set, the size of data window, search range；
Displacement Estimation module is used for the number of the data window according to the user setting, is corresponded to for the distribution of each data window
Processing thread, and the second surpassed with described by the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere
The cross-correlation coefficient of acoustic signals obtains the corresponding Displacement Estimation value of each data window；
Estimation module is strained, for being filtered respectively to each displacement estimated value, obtains each displacement estimated value difference
Corresponding strain estimated value；
Image-forming module obtains the elastogram result of the tissue for carrying out imaging to each strain estimated value.
Tissue elasticity imaging method provided by the invention and graphics processor receive tissue in image processor and generate shape
After the first ultrasonic signal and the second ultrasonic signal before and after becoming, triggering passes through to first ultrasonic signal and the second ultrasound
The processing of wave signal with obtain tissue Displacement Estimation result and strain estimated result, with according to strain estimated result carry out at
Picture obtains the process of the elastic image of tissue.The correlation of cross correlation algorithm used in the Displacement Estimation for obtaining user setting
After parameter, according to the quantity of data window, corresponding multiple processing threads are distributed, each thread that handles corresponds to a data window,
To, by multiple processing thread parallel cross-correlation calculations processing, with obtain the corresponding Displacement Estimation value of each data window to get
It is woven in different depth different moments corresponding Displacement Estimation value to group, and then lineation journey is obtained by way of filtering respectively again everywhere
Displacement Estimation is worth corresponding strain estimated value, to carry out imaging to each strain estimated value, obtain tissue it is elastic at
As result.According to the number of data window used in cross-correlation calculation, to distribute lineation journey everywhere in corresponding, at multi-threaded parallel
The mode of reason, while ensureing tissue elasticity imaging results accuracy, hence it is evident that improve the processing speed of elastogram.
Description of the drawings
Fig. 1 is the flow chart of tissue elasticity imaging method embodiment one of the present invention；
Fig. 2 is the flow chart of tissue elasticity imaging method embodiment two of the present invention；
Fig. 3 is the schematic diagram of graphics processor embodiment one of the present invention；
Fig. 4 is the schematic diagram of graphics processor embodiment two of the present invention.
Specific implementation mode
Fig. 1 is the flow chart of tissue elasticity imaging method embodiment one of the present invention, and in the present embodiment, which is imaged
Method is executed by graphics processor (Graphic Processing Unit, abbreviation GPU), and GPU settings are set in elastomeric check
In standby, as shown in Figure 1, the tissue elasticity imaging method includes：
Step 101, reception tissue generate the first ultrasonic signal before deformation and second after the tissue generation deformation
Ultrasonic signal.
Step 102 deposits the first ultrasonic signal, the second ultrasonic signal and the pre-set mutual related parameter of user
Enter into the corresponding global storages of GPU.
During carrying out elastogram to tissue, need to excite shearing wave in the tissue, for example pass through mechanical oscillation
Mode inspire shearing wave in the tissue.In order to measure the propagation characteristic of shearing wave in the tissue, to obtain the elasticity of tissue
Parameter to tissue emissions ultrasonic signal, that is, above-mentioned first ultrasonic signal before first shearing wave excitation, and receives ultrasonic echo letter
Number i.e. above-mentioned second ultrasonic signal, by the analyzing processing to the first ultrasonic signal and the second ultrasonic signal, to obtain
Characterize the parameters such as displacement, the strain of tissue elasticity deformation.Wherein, it generates the first ultrasonic signal and receives the second ultrasonic signal
Can be ultrasonic transducer, which can give the first ultrasonic signal and the second ultrasound signal transmission
GPU, so that it carries out subsequent processing.
For GPU after receiving above-mentioned first ultrasonic signal and the second ultrasonic signal, first ultrasonic wave is believed in triggering
Number and the second ultrasonic signal analyzing processing process, first, GPU by the first ultrasonic signal and the second ultrasonic signal and
The relevant parameter involved by cross correlation algorithm used in Displacement Estimation is carried out to be deposited into global storage.
In the present embodiment, it is normalized crosscorrelation algorithm to carry out cross correlation algorithm used by Displacement Estimation comprising such as
Under cross-correlation coefficient：
Wherein, x (t) is first ultrasonic signal, and y (t+ τ) is second ultrasonic signal, and R (u, τ) is described
The cross-correlation coefficient of first ultrasonic signal and second ultrasonic signal, t are sampled point, and u is the starting of the data window
Position, T are the size of the data window, and τ is described search range when calculating cross correlation value.
Wherein, above-mentioned described mutual related parameter includes the initial position u, the size T of data window, search range of data window
τ。
When cross-correlation coefficient R obtains maximum value, that is, the best match displacement of two ultrasonic signal data windows is found, i.e.,
The Displacement Estimation value of the corresponding tissue of a data window is obtained.
Step 103, the number according to the data window of the user setting distribute corresponding processing line for each data window
Journey, and pass through the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere and second ultrasonic signal
Cross-correlation coefficient, obtain the corresponding Displacement Estimation value of each data window.
In the present embodiment, the analyzing processing process to above-mentioned first ultrasonic signal and the second ultrasonic signal is into line position
It moves estimation and strains the process of estimation by the way of parallel processing.
Specifically, the computing cross-correlation of multiple data windows is carried out to the first ultrasonic signal and the second ultrasonic signal,
The number of specific data window can be preset.It is thus possible to determine required processing according to the number of the data window of setting
The quantity of thread, each computing cross-correlation for handling thread and corresponding to a data window.
To which GPU is after the initial position of number, each data window that data window is determined, size, for each processing
Thread distributes corresponding data window, so that the first ultrasonic signal in the reason thread parallel data window responsible to its and the everywhere
Two ultrasonic signals carry out computing cross-correlation as above.So that it is determined that coordinate corresponding when cross-correlation coefficient maximum value is pair
Answer the Displacement Estimation value of data window.
Step 104 is filtered each displacement estimated value respectively, obtains the corresponding strain of each displacement estimated value
Estimated value.
Step 105 carries out imaging to each strain estimated value, obtains the elastogram result of the tissue.
In the present embodiment, it is based on Elasticity correlation theory, each displacement estimated value can be carried out by way of filtering
It is filtered, you can obtain corresponding each strain estimated value, and then use grey scale mapping or color mapped, by each strain estimated value
It is mapped as corresponding gray level image or coloured image.
In the present embodiment, receives the first ultrasonic signal before and after tissue generation deformation in image processor and the second surpass
After acoustic signals, trigger by being estimated with the displacement for obtaining tissue to the processing of first ultrasonic signal and the second ultrasonic signal
Result and strain estimated result are counted, to be imaged according to strain estimated result, obtains the process of the elastic image of tissue.It is obtaining
After obtaining the relevant parameter of cross correlation algorithm used in the Displacement Estimation of user setting, according to the quantity of data window, distribution pair
The multiple processing threads answered, each thread that handles corresponds to a data window, to pass through multiple processing thread parallel cross-correlation meters
Calculation is handled, and different depth different moments corresponding position is woven in get to group to obtain the corresponding Displacement Estimation value of each data window
Estimated value is moved, and then lineation journey obtains the corresponding strain estimated value of each displacement estimated value by way of filtering again everywhere,
To carry out imaging to each strain estimated value, the elastogram result of tissue is obtained.According to data window used in cross-correlation calculation
Number, with distribution it is corresponding everywhere in lineation journey ensureing tissue elasticity imaging results by way of multi-threading parallel process
While accuracy, hence it is evident that improve the processing speed of elastogram.
Fig. 2 is the flow chart of tissue elasticity imaging method embodiment two of the present invention, as shown in Fig. 2, this method includes as follows
Step：
Step 201, reception tissue generate the first ultrasonic signal before deformation and second after the tissue generation deformation
Ultrasonic signal.
Step 202, by first ultrasonic signal, second ultrasonic signal and user it is pre-set mutually
Related parameter is deposited into the corresponding global storages of graphics processor GPU.
Step 203, that the initial position of the data window is read into from the global storage to the GPU is corresponding total
It enjoys in memory, the size of the data window and described search range is read into described GPU pairs from the global storage
In the register answered.
In the present embodiment, in order to further increase the processing speed of elastogram, carried out from data storage, the angle read
Following optimization：
Since GPU is corresponded to there are many different types of memory, global storage has maximum memory capacity, still
Its access speed is relatively slow, therefore, in order to improve the elastogram processing speed of GPU, in the present embodiment, with shared storage
Device, register improve the bandwidth of global storage as the buffering of global storage.
Specifically, the start position data u of data window is read in shared memory from global storage, to play buffering
Effect, improves the bandwidth of global storage.To data window size T, search range τ, under the premise of register number allows, from
It is read into register in global storage, to improve the access speed to these data.To shared memory and register
The method of salary distribution mainly considers the demand angle of access speed from size of data, data.
Step 204 carries out transposition operation respectively to the first ultrasonic signal and the second ultrasonic signal.
In general, the first ultrasonic signal and the second ultrasonic signal that GPU is received are connect in the form of column vector
It receives, for the ease of subsequent parallel processing, and in order to improve the reading speed of signal data, in the present embodiment, by turning
Operation is set, the first ultrasonic signal and the second ultrasonic signal are changed into the form of row vector.
Step 205 passes through the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere and described the
The maximum value of the cross-correlation coefficient of two ultrasonic signals.
Step 206 in the preset range of corresponding first coordinate points, is carried out when each cross-correlation coefficient is maximized
Fitting of a polynomial determines corresponding second coordinate points when the multinomial is maximized.
Step 207, according to each first coordinate points and corresponding each second coordinate points, determine that each displacement is estimated
Evaluation.
Since actual ultrasonic signal is discrete, in the case where sample frequency determines, obtained tissue
Deformation has error, and error is ± 0.5 sampling period.In order to reduce the error of Displacement Estimation, the mistake in Displacement Estimation is needed
Numerical interpolation is carried out in journey.The present embodiment uses the interpolation method of fitting of a polynomial, is fitted near cross-correlation coefficient maximum value
Polynomial equation, when i.e. above-mentioned second coordinate points of coordinate and cross-correlation coefficient when being maximized according to multinomial are maximized
Coordinate, that is, above-mentioned first coordinate points, can obtain more accurate tissue deformation, i.e. displacement estimated value.In simple terms, above-mentioned two
The difference of a coordinate points may be considered corresponding Displacement Estimation value.
Step 208 carries out median filter process respectively to each displacement estimated value.
Step 209 carries out SG filtering respectively to each displacement estimated value after the median filter process, obtains described
Each strain estimated value.
In the present embodiment, in order to ensure the accurate of Displacement Estimation result, and the accurate of estimated result is strained, it can be to each
Displacement Estimation value carries out median filter process, with filtering interference signals.
When carrying out strain estimation, SG filtering modes are used in the present embodiment.Specifically, Filtering Template a [n] is first determined,
Press again formula y (n)=a (0) * x (n)+a (2) * x (n-1)+...+a (n-1) * x (1) carry out difference, after x () is Displacement Estimation
Data, y (n) are that x (n) carries out SG filtering as a result, the strain that i.e. data of Displacement Estimation obtained after above-mentioned SG filtering is estimated
Evaluation.Wherein, n is the number of data window, and Filtering Template a [n] can experience setting in advance.
Each displacement estimated value and each strain estimated value are carried out transposition operation by step 210.
Since the first ultrasonic signal of front pair and the second ultrasonic signal have carried out transposition processing, for the ease of subsequently at
Conveniently as processing, meet the data imaging mode of imaging process, after obtaining each displacement estimated value and each strain estimated value, need
Transposition operation is carried out to each displacement estimated value and each strain estimated value.
Step 211 carries out imaging to each strain estimated value, obtains the elastogram result of tissue.
In the present embodiment, when carrying out Displacement Estimation and strain estimation, according to the number of data window used in cross-correlation calculation,
By way of multi-threading parallel process, while ensureing tissue elasticity imaging results accuracy, hence it is evident that improve elasticity at
The processing speed of picture；In addition, passing through the optimization processing to global storage so that the access speed of relevant parameter is also carried
Height further improves the processing speed of elastogram.
Fig. 3 is the schematic diagram of graphics processor embodiment one of the present invention, as shown in figure 3, the elastomeric check equipment includes：It connects
Receive module 11, storage processing module 12, Displacement Estimation module 13, strain estimation module 14, image-forming module 15.
Receiving module 11, for receiving, the tissue generates the first ultrasonic signal before deformation and the tissue generates institute
State the second ultrasonic signal after deformation.
Processing module 12 is stored, for first ultrasonic signal, second ultrasonic signal and user is pre-
The mutual related parameter being first arranged is deposited into the corresponding global storages of GPU, wherein the mutual related parameter is for into line position
The parameter for including in cross correlation algorithm used by estimation is moved, the mutual related parameter includes the initial position of data window, data
The size of window, search range.
Displacement Estimation module 13 is used for the number of the data window according to the user setting, for the distribution pair of each data window
The processing thread answered, and pass through the first ultrasonic signal and described second described in lineation journey parallel computation corresponding data window everywhere
The cross-correlation coefficient of ultrasonic signal obtains the corresponding Displacement Estimation value of each data window.
Estimation module 14 is strained, for being filtered respectively to each displacement estimated value, obtains each displacement estimated value point
Not corresponding strain estimated value.
Image-forming module 15 obtains the elastogram result of the tissue for carrying out imaging to each strain estimated value.
Wherein, the cross correlation algorithm includes：
Wherein, x (t) is first ultrasonic signal, and y (t+ τ) is second ultrasonic signal, and R (u, τ) is described
The cross-correlation coefficient of first ultrasonic signal and second ultrasonic signal, t are sampled point, and u is the starting of the data window
Position, T are the size of the data window, and τ is described search range when calculating cross correlation value.
The graphics processor of the present embodiment can be used for executing the technical solution of embodiment of the method shown in figure 1 above, in fact
Existing principle is similar with technique effect, and details are not described herein again.
Fig. 4 is the schematic diagram of graphics processor embodiment two of the present invention, as shown in figure 4, embodiment shown in Fig. 3 is basic
On, the storage processing module 12 is additionally operable to：The initial position of the data window is read into institute from the global storage
It states in the corresponding shared memories of GPU, the size of the data window and described search range is read from the global storage
Enter into the corresponding registers of the GPU.
The graphics processor further includes：First transposition module 21, the second transposition module 22.
First transposition module 21, for turning respectively to first ultrasonic signal and second ultrasonic signal
Set operation.
Second transposition module 22, for each displacement estimated value and each strain estimated value to be carried out transposition operation.
Further, the Displacement Estimation module 13, including：First computing unit 131, the second computing unit 132, third
Computing unit 133.
First computing unit 131, for passing through the first ultrasonic wave described in lineation journey parallel computation corresponding data window everywhere
The maximum value of signal and the cross-correlation coefficient of second ultrasonic signal；
Second computing unit 132, for when each cross-correlation coefficient is maximized corresponding first coordinate points it is pre-
If in range, carrying out fitting of a polynomial, determining corresponding second coordinate points when the multinomial is maximized；
Third computing unit 133, for according to each first coordinate points and corresponding each second coordinate points, determining
Each displacement estimated value.
Further, the strain estimation module 14 includes：First filter unit 141, the second filter unit 142.
First filter unit 141, for carrying out median filter process respectively to each displacement estimated value.
Second filter unit 142, for carrying out SG respectively to each displacement estimated value after the median filter process
Filtering, obtains each strain estimated value.
The graphics processor of the present embodiment can be used for executing the technical solution of embodiment of the method shown in figure 2 above, in fact
Existing principle is similar with technique effect, and details are not described herein again.
One of ordinary skill in the art will appreciate that：Realize that all or part of step of above method embodiment can pass through
The relevant hardware of program instruction is completed, and program above-mentioned can be stored in a computer read/write memory medium, the program
When being executed, step including the steps of the foregoing method embodiments is executed；And storage medium above-mentioned includes：ROM, RAM, magnetic disc or light
The various media that can store program code such as disk.
Finally it should be noted that：The above embodiments are only used to illustrate the technical solution of the present invention., rather than its limitations；To the greatest extent
Present invention has been described in detail with reference to the aforementioned embodiments for pipe, it will be understood by those of ordinary skill in the art that：Its according to
So can with technical scheme described in the above embodiments is modified, either to which part or all technical features into
Row equivalent replacement；And these modifications or replacements, various embodiments of the present invention technology that it does not separate the essence of the corresponding technical solution
The range of scheme.
Claims (8)
1. a kind of tissue elasticity imaging method, which is characterized in that including：
Step 1：Receive the tissue generate the first ultrasonic signal before deformation and the tissue generate after the deformation the
Two ultrasonic signals；
Step 2：By first ultrasonic signal, second ultrasonic signal and the pre-set cross-correlation ginseng of user
Number is deposited into the corresponding global storages of graphics processor GPU, wherein the mutual related parameter is for carrying out Displacement Estimation
The parameter for including in used cross correlation algorithm, the mutual related parameter include the initial position of data window, data window it is big
Small, search range；
Step 3：According to the number of the data window of the user setting, corresponding processing thread is distributed for each data window, and lead to
Cross the mutual of the first ultrasonic signal described in lineation journey parallel computation corresponding data window and second ultrasonic signal everywhere
Relationship number obtains the corresponding Displacement Estimation value of each data window；
Step 4：Each displacement estimated value is filtered respectively, obtains the corresponding strain estimation of each displacement estimated value
Value；
Step 5：Imaging is carried out to each strain estimated value, obtains the elastogram result of the tissue；
The step 3 includes：
Pass through the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere and second ultrasonic signal
Cross-correlation coefficient maximum value；
When each cross-correlation coefficient is maximized in the preset range of corresponding first coordinate points, fitting of a polynomial is carried out,
Determine corresponding second coordinate points when the multinomial is maximized；
According to each first coordinate points and corresponding each second coordinate points, each displacement estimated value is determined.
2. according to the method described in claim 1, it is characterized in that, the cross correlation algorithm includes following cross-correlation coefficient：
Wherein, x (t) is first ultrasonic signal, and y (t+ τ) is second ultrasonic signal, and R (u, τ) is described first
The cross-correlation coefficient of ultrasonic signal and second ultrasonic signal, t are sampled point, and u is the initial position of the data window,
T is the size of the data window, and τ is described search range when calculating cross-correlation coefficient.
3. according to the method described in claim 1, it is characterized in that, after the step 2, further include：
By the initial position of the data window from being read into the global storage in the corresponding shared memories of the GPU,
The size of the data window and described search range are read into the corresponding registers of the GPU from the global storage
In；
Transposition operation is carried out respectively to first ultrasonic signal and second ultrasonic signal；
Correspondingly, after the step 4, further include：
Each displacement estimated value and each strain estimated value are subjected to transposition operation.
4. according to the method described in claim 1, it is characterized in that, the step 4 includes：
Median filter process is carried out respectively to each displacement estimated value；
SG filtering is carried out respectively to each displacement estimated value after the median filter process, obtains each strain estimation
Value.
5. a kind of graphics processor, which is characterized in that including：
Receiving module, for receiving after tissue generates the first ultrasonic signal before deformation and the tissue generation deformation
Second ultrasonic signal；
Processing module is stored, for pre-setting first ultrasonic signal, second ultrasonic signal and user
Mutual related parameter be deposited into the corresponding global storages of graphics processor GPU, wherein the mutual related parameter be for into
The parameter for including in cross correlation algorithm used by row Displacement Estimation, the mutual related parameter include data window initial position,
The size of data window, search range；
Displacement Estimation module is used for the number of the data window according to the user setting, and corresponding place is distributed for each data window
Lineation journey, and pass through the first ultrasonic signal described in lineation journey parallel computation corresponding data window everywhere and second ultrasonic wave
The cross-correlation coefficient of signal obtains the corresponding Displacement Estimation value of each data window；
Strain estimation module obtains each displacement estimated value and corresponds to respectively for being filtered respectively to each displacement estimated value
Strain estimated value；
Image-forming module obtains the elastogram result of the tissue for carrying out imaging to each strain estimated value；
The Displacement Estimation module, including：
First computing unit, for passing through the first ultrasonic signal and institute described in lineation journey parallel computation corresponding data window everywhere
State the maximum value of the cross-correlation coefficient of the second ultrasonic signal；
Second computing unit, the preset range for corresponding first coordinate points when each cross-correlation coefficient is maximized
It is interior, fitting of a polynomial is carried out, determines corresponding second coordinate points when the multinomial is maximized；
Third computing unit, for according to each first coordinate points and corresponding each second coordinate points, determining described each
Displacement Estimation value.
6. graphics processor according to claim 5, which is characterized in that the cross correlation algorithm includes following cross correlation
Number：
Wherein, x (t) is first ultrasonic signal, and y (t+ τ) is second ultrasonic signal, and R (u, τ) is described first
The cross-correlation coefficient of ultrasonic signal and second ultrasonic signal, t are sampled point, and u is the initial position of the data window,
T is the size of the data window, and τ is described search range when calculating cross-correlation coefficient.
7. graphics processor according to claim 5, which is characterized in that the storage processing module is additionally operable to：It will be described
The initial position of data window in the global storage from being read into the corresponding shared memories of the GPU, by the data
The size and described search range of window in the global storage from being read into the corresponding registers of the GPU；
The graphics processor further includes：
First transposition module, for carrying out transposition fortune respectively to first ultrasonic signal and second ultrasonic signal
It calculates；
Second transposition module, for each displacement estimated value and each strain estimated value to be carried out transposition operation.
8. graphics processor according to claim 5, which is characterized in that the strain estimation module includes：
First filter unit, for carrying out median filter process respectively to each displacement estimated value；
Second filter unit is obtained for carrying out SG filtering respectively to each displacement estimated value after the median filter process
To each strain estimated value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510790828.6A CN105310727B (en) | 2015-11-16 | 2015-11-16 | Tissue elasticity imaging method and graphics processor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510790828.6A CN105310727B (en) | 2015-11-16 | 2015-11-16 | Tissue elasticity imaging method and graphics processor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105310727A CN105310727A (en) | 2016-02-10 |
| CN105310727B true CN105310727B (en) | 2018-07-17 |
Family
ID=55239724
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510790828.6A Active CN105310727B (en) | 2015-11-16 | 2015-11-16 | Tissue elasticity imaging method and graphics processor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105310727B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106805997B (en) * | 2016-12-26 | 2020-08-07 | 乐普（北京）医疗器械股份有限公司 | Elastic imaging method and device |
| WO2020061770A1 (en) * | 2018-09-25 | 2020-04-02 | 深圳迈瑞生物医疗电子股份有限公司 | Elastography method and device, and storage medium |
| CN111399039B (en) * | 2019-01-02 | 2023-04-07 | 无锡海斯凯尔医学技术有限公司 | Slope parameter extraction method, device and storage medium |
| CN112904349A (en) * | 2021-01-28 | 2021-06-04 | 清华大学 | Ultrasonic elastography method |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101530333A (en) * | 2002-07-31 | 2009-09-16 | 株式会社日立医药 | Ultrasonographic system, distortion distribution display method, and elastic modulus distribution display method |
| CN103857343A (en) * | 2011-08-19 | 2014-06-11 | 不列颠哥伦比亚大学 | Elastography using ultrasound imaging of a thin volume |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5535575B2 (en) * | 2009-10-27 | 2014-07-02 | ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー | Ultrasonic diagnostic equipment |
| JP5501813B2 (en) * | 2010-03-17 | 2014-05-28 | ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー | Ultrasonic diagnostic apparatus and control program therefor |
-
2015
- 2015-11-16 CN CN201510790828.6A patent/CN105310727B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101530333A (en) * | 2002-07-31 | 2009-09-16 | 株式会社日立医药 | Ultrasonographic system, distortion distribution display method, and elastic modulus distribution display method |
| CN103857343A (en) * | 2011-08-19 | 2014-06-11 | 不列颠哥伦比亚大学 | Elastography using ultrasound imaging of a thin volume |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105310727A (en) | 2016-02-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US5795296A (en) | Pipeline process for automatically measuring object boundary from ultrasound image samples | |
| DK2769241T3 (en) | Estimation and display to the vector Doppler imaging using transmissions with plane wave | |
| CN105310727B (en) | Tissue elasticity imaging method and graphics processor | |
| KR102025328B1 (en) | Apparatus and method for generating ultrasonic vector doppler image using plane wave synthesis | |
| CN102641137B (en) | Measurement of viscoelastic property using amplitude-phase modulation | |
| Rosenzweig et al. | GPU-based real-time small displacement estimation with ultrasound | |
| CN105266849B (en) | Real-time ultrasound elastograph imaging method and system | |
| Alessandrini et al. | Simulation of realistic echocardiographic sequences for ground-truth validation of motion estimation | |
| CN106659471A (en) | Autocorrelation guided cross-correlation in ultrasound shear wave elastography | |
| CN109875606B (en) | Ultrasonic CT sound velocity imaging method based on prior reflection imaging | |
| JP2013192643A (en) | Fluid velocity detection device and program | |
| CN101396281A (en) | Scanning transform method and device for color ultrasound image-forming | |
| US9286691B2 (en) | Motion estimation and compensation of feature-motion decorrelation | |
| CN104995530B (en) | Subject information acquisition device and subject information acquisition method | |
| CN108186050B (en) | Doppler blood flow velocity imaging method and system based on ultrasonic channel data | |
| US11953593B2 (en) | Systems, methods, and computer readable media for processing and compounding ultrasound images in the presence of motion | |
| US20230134503A1 (en) | Systems and methods for non-invasive pressure measurements | |
| US20130158403A1 (en) | Method for Obtaining a Three-Dimensional Velocity Measurement of a Tissue | |
| Wang et al. | An easily-achieved time-domain beamformer for ultrafast ultrasound imaging based on compressive sensing | |
| CN110477948B (en) | Elasticity imaging method and device, imaging equipment and storage medium | |
| Montero | New developments on quantitative imaging using ultrasonic waves | |
| US20100130862A1 (en) | Providing Volume Information On A Periodically Moving Target Object In An Ultrasound System | |
| CN106108942A (en) | Parallel ARFI formation method based on OpenCL | |
| CN110780302A (en) | Echo signal generation method based on continuous sound beam synthetic aperture | |
| CN104367346A (en) | Echo displacement detecting method and imaging method based on sound radiation force |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201221 Address after: Room 102, enlightenment star, 14 Donghai Road, Shizhong District, Zaozhuang City, Shandong Province, 277100 Patentee after: Haskell (Shandong) Medical Technology Co.,Ltd. Address before: Room B401, 530 building, University Science Park, Taihu International Science Park, Wuxi New District, Jiangsu Province Patentee before: Wuxi Hisky Medical Technologies Co.,Ltd. |
|
| TR01 | Transfer of patent right |