Ultrasonic super-resolution blood flow imaging method

Abstract

According to the ultrasonic super-resolution blood flow imaging method, super-resolution blood flow imaging is achieved through high-frequency components of ultrasonic echo signals on the basis of coupling of a preset structure sound field and contrast enhancement of an ultrasonic contrast agent. A current ultrasonic positioning microscope has the problem that the collection time is too long, and the super-resolution performance of the current ultrasonic positioning microscope can be influenced by depth. Structural sound field ultrasonic super-resolution is innovatively proposed, and the method comprises the following steps: firstly, based on a manually set target sound field, obtaining an ultrasonic pulse emission sequence corresponding to a specific structural sound field by adopting a phase inversion technology; secondly, an ultrasonic contrast agent serves as a scatterer in blood flow and is mutually coupled with the structural sound field to provide an echo signal; and finally, reducing aliasing high-frequency components in the echo signal through a deconvolution means to finally obtain a super-resolution image. According to the technology, a feasible scheme is provided for ultrahigh-resolution ultrasonic imaging, the limitation of traditional ultrasound in the aspects of instrument detection, biomedicine and the like can be broken through, and the application field of ultrasonic imaging is expanded.

Description

Technical Field

[0001] The present invention relates to an ultrasonic super-resolution blood flow imaging method, and in particular to an ultrasonic super-resolution method with customizable sound field design and contrast enhancement. Background Art

[0002] Ultrasound is a real-time, fast, and safe biomedical imaging method that has been widely used in clinical imaging and is of great significance for the evaluation of the physiological and pathological characteristics of various tissues in the human body. In recent years, contrast-enhanced ultrasound (CEUS) based on ultrasound contrast agents has received increasing attention. Microbubbles scatter in a nonlinear manner, thereby separating blood vessels and tissues, expanding the detection range of ultrasound. In response to the problem of low resolution of ultrasound imaging due to the objective diffraction limit, the idea of ​​optical super-resolution has been borrowed, and ultrasound localization microscopy (ULM) has emerged in recent years, which can break through the diffraction limit and obtain super-resolution blood flow imaging results.

[0003] However, ULM has the problem of long acquisition time. If the acquisition time is reduced, it is not easy to fully visualize the vascular structure, and its super-resolution performance will be affected by depth. In fact, the strategy of using structural sound field to improve spatial resolution is also a promising path. In other words, by analogy with the structured light field, structural sound field design can also be regarded as an encoder with great potential. Designing a spatially translated transmitting sound field with a fixed frequency sound field intensity can introduce prior information into the ultrasonic echo signal, and then restore the high-frequency information in the aliased signal through decoding in the reconstruction process.

[0004] It is worth noting that the current ultrasound imaging has the problem of being difficult to distinguish between the imaging target area and the background area, and lacks a truly manually customized regional scanning method. However, the two-dimensional or three-dimensional structural sound field can set the soundless field intensity for the background part that is not of interest, thereby avoiding the noise generated by the background tissue echo and improving the signal-to-noise ratio of the ultrasound image. In addition, the premise of structural sound field super-resolution requires sufficient ultrasound information identification, so combining contrast-enhanced ultrasound (CEUS) can go a step further and better improve the resolution of scatterers. Summary of the invention

[0005] In view of the above problems and the deficiencies of the prior art, the present invention discloses a customizable sound field design contrast enhancement ultrasound super-resolution method.

[0006] In order to solve the problem of long acquisition time, the present invention innovatively proposes an ultrasonic super-resolution method under the design of structural sound field, and combines it with ultrasound contrast imaging super-resolution to achieve an even more powerful effect. Specifically, it consists of the following three innovative parts: First, based on the artificially set target sound field intensity distribution, the phase inversion technology is used to design the ultrasonic pulse emission sequence corresponding to the specific structural sound field; second, the ultrasonic contrast agent acts as a scatterer in the blood flow, and couples with the structural sound field to provide an echo signal; finally, through deconvolution methods such as phase shift superposition, the aliased high-frequency components in the echo signal are restored to reconstruct the super-resolution image.

[0007] Inspired by the super-resolution idea in the field of optics, the present invention proposes a scheme to form an intensity distribution with a specific spatial frequency, or a distribution with self-similar characteristics, or a random distribution that satisfies the RIP condition during the ultrasonic sound field design stage, and then use the structured sound field pattern sequence transmission and reception combined with ultrasonic contrast imaging to introduce high-frequency components in the frequency domain for super-resolution imaging. The intensity distribution of a specific spatial frequency can introduce high-frequency information It is worth noting that the design of the sound field can only work when the signal-to-noise ratio and signal-to-background ratio are good, that is, the premise of structural sound field super-resolution requires sufficient ultrasonic information to be identified. In conventional ultrasound B-mode images, the signal-to-noise ratio is often very low, and structural sound field super-resolution imaging is not very helpful. However, ultrasound contrast imaging introduces contrast agents, which enables ultrasound to sensitively image microvessels. The nonlinear scattering of microbubbles increases the contrast of tiny blood vessels, providing unprecedented enough information for the structural sound field to further achieve super-resolution. Therefore, the structural sound field "illumination" based on contrast-enhanced ultrasound (CEUS) can have an even more powerful effect.

[0008] In summary, the present invention reduces the half-height full width of the scatterer by designing the parameters of the pulse sequence emitted by each array element corresponding to the structural sound field, thereby overcoming the diffraction limit, achieving super-resolution, and being able to generate a novel ultrasonic array pulse sequence corresponding to the two-dimensional xz plane sound field. In addition to the innovation of the above overall framework structure, a specific new optimization strategy is also proposed, including the innovative integration of Wiener deconvolution, total variance, RL and other optimization methods in the reconstruction of the structural sound field image; it also includes a new quantitative evaluation index for the construction accuracy of the structural sound field, and defines the evaluation index according to the normalized matching coefficient between the target sound field intensity and the actual sound field intensity; at the same time, the Gerchberg-saxton algorithm is also improved, that is, in view of the problem that the Gerchberg-saxton algorithm has low robustness due to the initial random value, the total variation is introduced, and the sound field intensity matching coefficient and the sound field smoothness are considered, so as to solve the ultrasonic emission parameters more quickly and robustly. The same method is also applicable to the design of the pulse sequence parameters of the matrix probe ultrasonic array corresponding to the preset three-dimensional structural sound field.

[0009] The technical route disclosed in the present invention provides a practical solution for ultra-high resolution ultrasonic imaging, and can break through the limitations of traditional ultrasound in instrument detection, biomedicine, energy focusing, etc., and expand the application field of ultrasonic imaging. At the same time, the pulse sequence parameter generation algorithm corresponding to the two-dimensional and three-dimensional structural sound fields can also obtain a more accurate and customized ultrasonic imaging range, improve the signal-to-noise ratio of ultrasonic images, and open up a prospect of computational imaging with more adaptive characteristics. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Figure 1 : A schematic block diagram of the overall process of the super-resolution algorithm described in the present invention; Figure 2 : The result diagram of the structural sound field design of the present invention; Implementation

[0011] The structural sound field ultrasound contrast super-resolution imaging is further described below through specific embodiments to facilitate a better understanding of the present invention. However, any changes or substitutions that can be easily conceived by a technician familiar with the technical field within the technical scope disclosed in the present invention should be covered within the scope of protection of the claims of the present invention.

[0012] First combine Figure 1 The overall process of the super-resolution algorithm proposed in the present invention is described in detail, and the specific steps are as follows: (1) Set the sound field intensity distribution of N preset structures for ultrasonic scanning to S i (i=1,2,...,N), where N>2; (2) According to the preset structural sound field intensity distribution S i , using the phase inversion technique to obtain the probe array element parameters corresponding to the current i-th ultrasonic pulse transmission sequence, including the phase delay t i and apodization i ; (3) Collect echo signals R in sequence i , and the imaging result I is obtained based on the classical beamforming raw ; (4) Repeat the above steps (2)-(3) to obtain all N corresponding imaging results I raw i (i=1,2,...,N); (5) Solve the inverse problem based on all N corresponding imaging results and illumination distribution to obtain the preliminary super-resolution result I of the structural sound field SISR ; (6) For the preliminary super-resolution results I SISR , the super-resolution blood flow image I is finally obtained through the SVD filtering method f .

[0013] Then, in order to verify the effectiveness of the sound field design, the present invention tried the rationality and effectiveness of the structural sound field design on the ordinary linear array. The specific sound field results are as follows: Figure 2 shown.

[0014] An embodiment of the present invention adopts the Verasonics Vantage256 ultrasonic operating system, and the probe used is the L11_5v probe with a center frequency of 7.81 MHz. A variety of quantitative evaluation indicators of structural sound fields are designed and the Gerchberg-saxton algorithm is optimized to generate a sound field intensity distribution with a spatial frequency of 1.8 mm at a depth of 18 mm.

[0015] In the process of sound field design, considering the problem that the classic Gerchberg-saxton algorithm has low robustness due to the initial random value, the present invention introduces a constrained optimization improvement algorithm in the iterative process. Specifically, in the process of iteratively solving the phase delay and apodization in the focal plane and the ultrasonic transducer plane, different from the direct replacement of the Gerchberg-saxton algorithm, the smoothness of the sound field intensity in the horizontal (x) and axial (z) directions is considered, and the following formula is set as the optimization target, while considering the consistency of the sound field intensity change before and after the iteration to solve it.

[0016]

[0017] Where x and z refer to the horizontal and axial positions respectively, S is the sound field intensity, i is the number of iterations, f represents the relationship between the transmitted sound field and the ultrasonic transmission parameters, a and t are the apodization and phase delay of the ultrasonic probe respectively, and p is a weight parameter used to weigh the smoothness and intensity matching of the sound field.

[0018] like Figure 2 As shown, one embodiment of the present invention designs a sound field with a fixed spatial frequency, wherein the horizontal axis is the horizontal direction, the vertical axis is the depth direction, the light and dark colors represent the intensity, the structured sound field intensity is clearly discernible, and the spatial frequency is 1.8 mm, which is consistent with the preset structural sound field, which fully demonstrates the effectiveness of the sound field design strategy provided by the present invention.

[0019] exist Figure 2 Based on the design of the structural sound field, the mature super-resolution ideas in structured light are borrowed. After collecting echo signals for multiple times, the algorithm for solving the inverse problem is used, and the phase shift superposition method is preferably used to obtain a finer spatial resolution. Therefore, it has advantages over traditional plane wave B-mode imaging. At the same time, the area with a low sound field intensity also shows low signal characteristics after imaging, which can meet the requirements of ultrasound custom scanning of specific areas and accurately separate the target area and background area.

Claims

1. An ultrasonic super-resolution blood flow imaging method, characterized in that: Based on the coupling of preset structural sound field and contrast enhancement of ultrasonic contrast agent, super-resolution blood flow imaging is achieved by utilizing the high-frequency components of ultrasonic echo signals; wherein the preset structural sound field includes one-dimensional, two-dimensional and three-dimensional.

2. The ultrasonic super-resolution imaging method according to claim 1, characterized in that: The following steps are involved: (1) Set the sound field intensity distribution of N preset structures for ultrasonic scanning to S i (i=1,2,...,N), where N>2; (2) According to the preset structural sound field intensity distribution S i , using the phase inversion technique to obtain the probe array element parameters corresponding to the current i-th ultrasonic pulse transmission sequence, including the phase delay t i and apodization i ; (3) Collect echo signals R in sequence i , and the imaging result I is obtained based on the classical beamforming raw ; (4) Repeat the above steps (2)-(3) to obtain all N corresponding imaging results I raw i (i=1,2,...,N); (5) Solve the inverse problem based on all N corresponding imaging results and illumination distribution to obtain the preliminary super-resolution result I of the structural sound field SISR ; (6) For the preliminary super-resolution results I SISR , the super-resolution blood flow image I is finally obtained through the SVD filtering method f .

3. The preset structural sound field according to claim 1, characterized in that: The preset structural sound field is located in the focal plane, the sound field intensity has a distribution with a specific spatial frequency, and the phase interval between adjacent stripes is 0 to π; or it has a distribution with self-similar characteristics, or it has a random distribution that meets the RIP condition.

4. The phase inversion step according to claim 2, characterized in that: The structural sound field S required for each launch is estimated by using the improved GSA inverse solution or multi-focus method iterative fitting, or by using machine learning methods. i The corresponding ultrasonic transmission parameters include the phase delay t of each ultrasonic array element. i and apodization i (i=1,2,...,N).

5. The improved GSA algorithm as claimed in claim 4, characterized in that: By introducing total variation and considering the sound field intensity matching coefficient and sound field smoothness constraints, the optimization function preferably adopts the following formula:

6. Where x and z refer to the horizontal and axial positions respectively, S is the sound field intensity, i is the number of iterations, f represents the relationship between the transmitted sound field and the ultrasonic transmission parameters, a and t are the apodization and phase delay of the ultrasonic probe respectively, and p is a weight parameter used to weigh the smoothness and intensity matching of the sound field.

7. The ultrasonic emitting step according to claim 2, characterized in that: The transmitting system is compatible with linear array, convex array and matrix ultrasonic imaging probes; wherein the ultrasonic probe transmits an ultrasonic sequence with a central frequency of 1 to 30 MHz and 1 to 20 ultrasonic transmission cycles; the ultrasonic mechanical index MI is less than 1.9; and the structural sound field is provided by the transmitted ultrasonic sequence.

8. The structural sound field super-resolution step according to claim 2, characterized in that: According to the preset structure-borne sound field distribution S1 to S n , using matched filtering or sparse reconstruction or waveform inversion strategy, we can get I SISR .

Images