A full matrix ultrasonic fast imaging method suitable for multi-layer bone tissue

Abstract

The application discloses a full-matrix ultrasonic rapid imaging method suitable for multilayer bone tissue, and comprises the following steps: step one, building an experimental platform; step two, collecting wave field information of an imaging target by using a full-matrix capturing method; step three, obtaining the sound velocity distribution in the wave field; step four, completing the depth extrapolation of single-angle frequency domain wave field; step five, perfecting the wave field information; and step six, obtaining final ultrasonic imaging. The full-matrix ultrasonic rapid imaging method suitable for multilayer bone tissue is designed, and high-precision and high-contrast multilayer structure bone imaging is realized while reducing the calculation burden.

Description

Technical Field

[0001] This invention relates to the field of full-matrix capture ultrasound imaging, and particularly to a full-matrix ultrasound rapid imaging method suitable for multi-layer bone tissue. Background Technology

[0002] For bone, a biological hard tissue with a complex multi-layered structure, traditional ultrasound imaging methods that are only applicable to a single layer with a fixed sound velocity are difficult to apply directly due to its significant changes in sound velocity, position-dependent anisotropy, and high attenuation. Moreover, most known ultrasound imaging methods that can achieve bone imaging have high computational complexity and cannot achieve real-time imaging in clinical practice.

[0003] By capturing ultrasonic signals from all ultrasonic elements during the reception of each transmission, full-matrix capture imaging can collect and record more dimensional target information, thus providing higher quality imaging results compared to traditional ultrasound imaging. However, early full-matrix capture ultrasound imaging was mostly based on time-domain data, which inevitably resulted in a significant computational burden.

[0004] Meanwhile, it should be noted that for hard biological tissues such as bone, due to the significant difference in sound velocity between bone and surrounding soft tissues, ultrasound waves experience significant reflection, scattering, and attenuation at their interfaces, hindering the propagation and collection of ultrasound signals and thus increasing the difficulty of imaging. Although existing full-matrix capture ultrasound imaging methods can image bone phantoms, their massive computational data requirements and complex calculation process prevent them from being directly applied in clinical settings to achieve real-time imaging.

[0005] Traditional full-matrix capture ultrasound imaging has the following limitations: it requires a known sound velocity distribution model. Without a prior sound velocity model, the recorded ultrasound signal is inevitably subject to interference, which inevitably leads to distortion and deformation of the imaging results; the computational data volume is enormous. Traditional full-matrix capture ultrasound imaging directly utilizes ultrasound data recorded in the time domain, resulting in a larger data storage and processing volume, making it unsuitable for ultrasound platforms with limited computing resources; the significant attenuation of ultrasound signals at the bone-soft tissue boundary severely affects the final imaging quality. The large difference in acoustic impedance between bone and soft tissue prevents ultrasound waves from effectively propagating into the bone, making it difficult to achieve multi-layered, irregular bone imaging.

[0006] Therefore, there is an urgent need to design a more efficient and practical imaging method for ultrasound imaging of bones. Summary of the Invention

[0007] To overcome the shortcomings of existing technologies, this invention provides a full-matrix ultrasound rapid imaging method suitable for multi-layer bone tissue, which reduces the computational burden while achieving high-precision, high-contrast multi-layer bone imaging.

[0008] To achieve the aforementioned objectives of the invention, the technical solution adopted to solve its technical problems is as follows:

[0009] A rapid full-matrix ultrasound imaging method applicable to multi-layer bone tissue includes the following steps:

[0010] Step 1: Set up the experimental platform;

[0011] Step 2: Collect wavefield information of the target to be imaged using the full matrix capture method;

[0012] Step 3: Obtain the sound velocity distribution in the wave field;

[0013] Step 4: Complete the extrapolation of the single-angle frequency domain wavefield at different depths;

[0014] Step 5: Improve wave field information;

[0015] Step 6: Obtain the final ultrasound image.

[0016] Furthermore, step one includes the following:

[0017] Two identical linear array transducers are placed face-to-face in a water tank. Each transducer has N array elements. The cortical bone imaging target to be imaged is placed between the two transducers, and water is injected into the water tank to immerse the imaging target and the ultrasound transducers.

[0018] Furthermore, step two includes the following:

[0019] Each element on a transducer sequentially transmits an ultrasonic pulse signal with a fixed center frequency. At this time, all elements on that transducer, along with all elements on another transducer, serve as receivers to record the sound field information. After all transducers complete the transmission-reception process, 2N sets of 2×N×N data are obtained. t The sound field data, where N t The number of samples.

[0020] Furthermore, step three includes the following:

[0021] Based on the difference between the theoretical earliest arrival time of ultrasound and the actual earliest arrival time, the following optimization objective function is obtained:

[0022]

[0023] Where S is the distribution of the slowness (reciprocal of velocity) to be estimated, T(S) is the theoretical earliest arrival time of the ultrasonic wave, and T obsThis refers to the earliest measured time of arrival of ultrasound, where J(S) represents the different propagation paths under the current slowness model. Based on Bayesian estimation, the final slowness distribution estimate is obtained, and taking its reciprocal yields the sound velocity distribution estimate.

[0024] Furthermore, step four includes the following:

[0025] The wavefield information in the time domain is converted to the frequency domain using a Fast Fourier Transform. Then, one ultrasonic transducer is selected, and a sound velocity estimation model is used. The sound velocity model is layered based on the ultrasonic frequency domain information emitted by the transducer, and frequency domain wave field extrapolation at different depths is performed based on the frequency domain wave field at the location of the transducer to complete the rapid reconstruction of the imaging target on one side of the currently selected transducer.

[0026] Furthermore, the time-domain ultrasonic field information p is transformed into the frequency domain using three dimensions to obtain P(ω,k). e ,k r ), where ω is the frequency, k e With k r Refers to the coordinate information relative to the transmitting and receiving array elements in the frequency domain space, and the velocity model It is vertically divided into 10 layers. Then, the frequency domain wave field at the location of one transducer is used as the reference z. e =0, perform wave field extrapolation, z e =z i-1 With z e The wave field relationship between 0 and 0 can be expressed by the following formula:

[0027]

[0028]

[0029] Among them, L m for The physical thickness of the velocity layer, where j is the imaginary unit and e is the base of the natural logarithm function, is determined using the Stolt mapping relation, which represents P(ω,k). e ,k r ) transformed into F T (k x ,k z ), where k x =k e +k r , After obtaining complete frequency domain wavefield data, for F T (k x ,k z Perform a two-dimensional inverse Fourier transform to obtain the scattering distribution at different positions on one side. The formula is:

[0030]

[0031] Here, (x,z) refers to the two-dimensional spatial coordinates of the imaging region.

[0032] Furthermore, step five includes the following:

[0033] Select an ultrasonic transducer in another direction and repeat step four to complete the rapid reconstruction on the other side.

[0034] Furthermore, step six includes the following:

[0035] By fusing the two unilaterally reconstructed images, information about the target from different orientations is improved, resulting in a final, complete reconstructed image.

[0036] By employing the above technical solutions, this invention has the following advantages and positive effects compared with the prior art:

[0037] This invention further improves upon the traditional full-matrix capture ultrasound imaging method. First, based on the ultrasonic wave propagation time, a Bayesian optimization estimation is performed on the sound velocity distribution model to obtain approximate sound velocity distribution information, effectively reducing the impact caused by unknown and large sound velocity variations. Based on this sound velocity model, a three-dimensional Fourier transform is performed on the transmitted and received wavefield information, which can significantly reduce the amount of data and greatly improve data storage and computation efficiency. At the same time, it also utilizes full-matrix information to ensure reliable imaging quality. Attached Figure Description

[0038] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are merely some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings:

[0039] Figure 1 This is a flowchart of a full-matrix rapid ultrasound imaging method applicable to multi-layer bone tissue in an embodiment of the present invention;

[0040] Figure 2 This is a schematic diagram of the signal transmission and acquisition system of the ultrasonic transducer in an embodiment of the present invention;

[0041] Figure 3 This is a longitudinal section image of the cortical bone to be imaged in this embodiment of the invention, obtained based on CT.

[0042] Figure 4 These are imaging results applied to cortical bone in embodiments of the present invention;

[0043] Figure 5 This is an imaging result diagram of a conventional synthetic aperture ultrasound imaging method based on ray tracing under the same conditions in an embodiment of the present invention. Detailed Implementation

[0044] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0045] This embodiment discloses a full-matrix rapid ultrasound imaging method applicable to multi-layer bone tissue, including the following steps:

[0046] Step 1: Set up the experimental platform;

[0047] Furthermore, step one includes the following:

[0048] Two identical linear array transducers are placed face-to-face in a water tank. Each transducer has N array elements. The cortical bone imaging target to be imaged is set between the two transducers. Water is then poured into the water tank to immerse the imaging target and the ultrasound transducers, thus completing the construction of the experimental platform for this ultrasound imaging method.

[0049] Step 2: Collect wavefield information of the target to be imaged using the full matrix capture method;

[0050] Furthermore, step two includes the following:

[0051] Each element on a transducer sequentially transmits an ultrasonic pulse signal with a fixed center frequency. At this time, all elements on that transducer, along with all elements on another transducer, serve as receivers to record the sound field information. After all transducers complete the transmission-reception process, 2N sets of 2×N×N data are obtained. t The sound field data, where N t The number of samples is [number]. Wavefield data acquisition was completed using the full matrix capture method, i.e., the "single-transmit-full-receive" approach, obtaining different positional information of the imaging target from multiple angles.

[0052] Step 3: Obtain the sound velocity distribution in the wave field;

[0053] Furthermore, step three includes the following:

[0054] Based on the difference between the theoretical earliest arrival time of ultrasound and the actual earliest arrival time, the following optimization objective function is obtained:

[0055]

[0056] Where S is the distribution of the slowness (reciprocal of velocity) to be estimated, T(S) is the theoretical earliest arrival time of the ultrasonic wave, and T obs This refers to the earliest measured time of arrival of ultrasound, where J(S) represents the different propagation paths under the current slowness model. Based on Bayesian estimation, the final slowness distribution estimate is obtained, and taking its reciprocal yields the sound velocity distribution estimate. This enables the estimation of sound velocity distribution by combining the first arrival wave method with Bayesian estimation, which not only facilitates data collection but also simplifies operation and greatly improves imaging efficiency.

[0057] Step 4: Complete the extrapolation of the single-angle frequency domain wavefield at different depths;

[0058] Furthermore, step four includes the following:

[0059] The wavefield information in the time domain is converted to the frequency domain using a Fast Fourier Transform. Then, one ultrasonic transducer is selected, and a sound velocity estimation model is used. Based on the ultrasonic frequency domain information emitted by the transducer, the sound velocity model is layered, and based on the frequency domain wavefield at the transducer's location, frequency domain wavefield extrapolation at different depths is performed to complete the rapid reconstruction of one side of the imaging target on the currently selected transducer side. By layering the frequency domain wavefield and then extrapolating layer by layer, not only is the amount of wavefield data computation reduced, but imaging at different depths is also achieved, which can meet different imaging requirements.

[0060] Step 5: Improve wave field information;

[0061] Furthermore, step five includes the following:

[0062] Select an ultrasonic transducer in another direction and repeat step four to complete the rapid reconstruction on the other side, thereby obtaining information about the imaging target from different orientations and improving imaging quality.

[0063] Step 6: Obtain the final ultrasound image.

[0064] Furthermore, step six includes the following:

[0065] By fusing the two unilaterally reconstructed images, information about the target from different orientations is improved, resulting in a final, complete reconstructed image.

[0066] Example:

[0067] Figure 2 This is a schematic diagram of the signal transmission and acquisition system of the dual linear array transducer in an embodiment of the present invention. Figure 2As shown, two identical linear array transducers were placed parallel to each other, each with 128 elements. The cortical bone phantom to be imaged was placed between the two transducers, and water was poured into a water tank to immerse the sample and the ultrasonic transducers. Each element of the transducer sequentially emitted an ultrasonic signal with a fixed center frequency of 3.5 MHz, the center-to-center distance between adjacent elements was 0.3 mm, and the receiving sampling rate was 25 MHz.

[0068] Figure 1 This is a flowchart illustrating a full-matrix rapid ultrasound imaging method applicable to multi-layered bone tissue in an embodiment of the present invention. Figure 1 As shown, the specific implementation steps of this embodiment are as follows:

[0069] Step 1: Place two identical linear array transducers in parallel alignment. Each linear array transducer has 128 array elements. Place the cortical bone phantom to be imaged between the two transducers and inject water into the water tank to immerse the sample and the ultrasonic transducers.

[0070] Step 2: Each element on the two transducers sequentially transmits an ultrasonic pulse signal with a center frequency of 3.5MHz. At the same time, all other elements on the two transducers act as receivers to record the ultrasonic signal. The final recorded result is p, with dimensions of 256×256×3000.

[0071] Step 3: Based on the recorded ultrasound signals, extract the first arrival time of different array elements and record it as T. obs The objective function is obtained by subtracting the theoretical earliest arrival time of ultrasound from the actual earliest arrival time:

[0072]

[0073] Where S is the distribution of the slowness (reciprocal of velocity) to be estimated, T(S) is the theoretical earliest arrival time of the ultrasonic wave, and T obs This represents the earliest measured time of arrival of the ultrasound, and J(S) refers to different propagation paths under the current slowness model. Based on Bayesian estimation, the final slowness distribution estimate is obtained, and taking its reciprocal yields the sound velocity distribution estimate.

[0074] Step 4: Transform the time-domain ultrasonic field information p into the frequency domain using three-dimensional transformation to obtain P(ω,k). e ,k r ), where ω is the frequency, k e With k r This refers to the coordinate information relative to the transmitting and receiving array elements in the frequency domain. (Regarding the velocity model) It is vertically divided into 10 layers. Then, the frequency domain wave field at the location of one transducer is used as the reference z. e=0, perform wavefield extrapolation, z e =z i-1 With z e The wave field relationship between 0 and 0 can be expressed by the following formula:

[0075]

[0076]

[0077] Among them, L m for The physical thickness of the velocity layer, where j is the imaginary unit and e is the base of the natural logarithm. Using the Stolt mapping, P(ω,k) e ,k r ) transformed into F T (k x ,k z ), where k x =k e +k r , After obtaining complete frequency domain wavefield data, for F T (k x ,k z Perform a two-dimensional inverse Fourier transform to obtain the scattering distribution at different positions on one side. The formula is:

[0078]

[0079] Here, (x,z) refers to the two-dimensional spatial coordinates of the imaging region.

[0080] Step 5: Repeat the operation in Step 4 on the wavefield information received from the other side to obtain the wavefield information from the other side.

[0081] Step Six: Place the two By fusing the images, we obtain the final, complete image result.

[0082] Figure 4 This is the imaging result of the cortical bone phantom according to an embodiment of the present invention. The imaging results show that the inner and outer boundaries of the cortical bone are clearly and accurately reconstructed, and the shape and position highly coincide with the real model, with a boundary error of only 9.3%. Figure 5The results are obtained by imaging the same cortical bone under the same conditions using a single transducer and traditional ray-tracing-based synthetic aperture ultrasound imaging method. It is evident that, within the same dynamic range, the traditional method can only visualize the upper surface of the upper cortical bone, failing to effectively image the lower surface. In contrast, the method used in this invention can visualize the complete upper and lower cortical bone surfaces. Therefore, the proposed spectral domain analysis-based full-matrix capture ultrasound imaging method demonstrates significantly superior imaging performance compared to traditional ray-tracing-based synthetic aperture ultrasound imaging algorithms.

[0083] In the embodiments, the ultrasound full-matrix capture imaging method based on spectral domain analysis of the present invention can achieve accurate and efficient imaging of cortical bone. Furthermore, compared with the traditional synthetic aperture method based on ray tracing, the present invention achieves better bone imaging results through plane wave emission, overcoming the shortcomings of traditional ultrasound imaging algorithms for bone tissue.

[0084] The above description is merely a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A full-matrix rapid ultrasound imaging method applicable to multi-layer bone tissue, characterized in that, Includes the following steps: Step 1: Set up the experimental platform; Step one includes the following: Two identical linear array transducers were placed face-to-face in a water tank, each transducer having... Each array element sets the cortical bone imaging target to be imaged between two transducers, and water is injected into the water tank to immerse the imaging target and the ultrasound transducers. Step 2: Collect wavefield information of the target to be imaged using the full matrix capture method; Step 3: Obtain the sound velocity distribution in the wave field; Step three includes the following: Based on the difference between the theoretical earliest arrival time of ultrasound and the actual earliest arrival time, the following optimization objective function is obtained: in, It is the distribution of slowness to be estimated. This is the theoretical earliest arrival time of ultrasound. It is the earliest actually measured time of ultrasound arrival. This refers to different propagation paths under the current slowness model; based on Bayesian estimation, the final slowness distribution estimate is obtained, and taking its reciprocal yields the sound speed distribution estimate. ; Step 4: Complete the extrapolation of the single-angle frequency domain wavefield at different depths; Step four includes the following: The wavefield information in the time domain is converted to the frequency domain using a Fast Fourier Transform. Then, one ultrasonic transducer is selected, and a sound velocity estimation model is used. The sound velocity model is layered based on the ultrasonic frequency domain information emitted by the transducer, and frequency domain wave field extrapolation at different depths is performed based on the frequency domain wave field at the location of the transducer to complete the rapid reconstruction of one side of the imaging target on the side of the currently selected transducer. Information of ultrasonic fields in the time domain By transforming from three dimensions to the frequency domain, we obtain ,in, It is a frequency variable. It is the spatial wavenumber of the transmitting element along the transmission direction. It is the spatial wavenumber of the receiving element along the radial direction of the transmitting element, which is used in the velocity model. It is vertically divided into 10 layers. Then, the frequency domain wave field at the location of one transducer was used as the reference. Perform wave field extrapolation. and The wavefield relationship between them can be expressed by the following formula: in, for The physical thickness of the velocity layer, It is the imaginary unit. It is the base of the natural logarithm function. Using the Stolt mapping, we can... Transform into ,in, , , and It refers to the spatial wavenumber along the x and z directions in a two-dimensional coordinate system. After obtaining complete frequency domain wavefield data, it is used for... Performing a two-dimensional inverse Fourier transform yields the scattering distribution at different locations on one side. The formula is: in, This refers to the two-dimensional spatial coordinates of the imaging area; Step 5: Improve wave field information; Step five includes the following: Select an ultrasonic transducer in another direction and repeat step four to complete the rapid reconstruction on the other side. Step Six: Obtain the final ultrasound image; Step six includes the following: By fusing the two unilaterally reconstructed images, information about the target from different orientations is improved, resulting in a final, complete reconstructed image.

2. The method for rapid full-matrix ultrasound imaging of multi-layered bone tissue according to claim 1, characterized in that, Step two includes the following: Each element on a transducer sequentially transmits an ultrasonic pulse signal with a fixed center frequency. At this time, all elements on that transducer, along with all elements on another transducer, serve as receivers recording the sound field information. After all transducers complete the transmission-reception process, the sound field information is obtained. Group dimension is The sound field data, among which, The number of samples.

Images