阅读说明：本技术 一种医学影像逆时偏移成像方法及装置 (Medical image reverse time migration imaging method and device ) 是由 佟小龙 张家豹 葛勇 马国栋 于 2020-12-18 设计创作，主要内容包括：本发明涉及医学成像领域,具体是涉及一种医学影像逆时偏移成像方法及装置,包括以下步骤获取模型参数,根据所述模型参数,确定时间延拓步长以及偏移孔径范围；在孔径范围内添加随机速度边界,生成孔径范围内的随机速度边界模型；确定最大接收时间,根据所述最大接收时间,获取反向传播最后两个时刻的正演震源波场；自最大接收时间,将单炮数据反向传播回地下；进行初步的医学成像；存储初步的医学成像结果；去除初步的医学成像结果的成像噪音；恢复成像结果的空间坐标位置信息,得到最终的医学成像结果,本发明可以实现低成本高分辨率的医学成像,可以以超声成像的硬件成本,实现核磁共振的成像精度。(The invention relates to the field of medical imaging, in particular to a method and a device for reverse-time migration imaging of medical images, which comprises the following steps of obtaining model parameters, and determining a time continuation step length and a migration aperture range according to the model parameters; adding a random speed boundary in the aperture range to generate a random speed boundary model in the aperture range; determining the maximum receiving time, and acquiring forward source wave fields of the last two moments of backward propagation according to the maximum receiving time; from the maximum receive time, back-propagating the single shot data to the subsurface; performing preliminary medical imaging; storing preliminary medical imaging results; removing imaging noise of a preliminary medical imaging result; the invention can realize low-cost and high-resolution medical imaging and realize the imaging precision of nuclear magnetic resonance with the hardware cost of ultrasonic imaging.)

1. A method for reverse time migration imaging of medical images, comprising the steps of:

obtaining model parameters, and determining a time continuation step length and an offset aperture range according to the model parameters;

adding a random speed boundary in the aperture range to generate a random speed boundary model in the aperture range;

determining the maximum receiving time, acquiring forward source wave fields of the last two moments of backward propagation according to the maximum receiving time, and determining the maximum receiving time through the time continuation step length;

from the maximum receive time, back-propagating the single shot data to the subsurface;

performing preliminary medical imaging;

storing preliminary medical imaging results;

removing imaging noise of a preliminary medical imaging result;

and recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

2. The method of claim 1, wherein the model parameters comprise a migration parameter and single shot data.

3. The method as claimed in claim 2, wherein a maximum receiving time is determined, and forward source wavefields backward propagating for the last two moments are obtained according to the maximum receiving time, and the method comprises the following steps:

generating a seismic source wavelet;

forward modeling a seismic source wave field to the maximum receiving time by using a finite difference method according to the seismic source wavelet and the random velocity boundary model;

and acquiring forward seismic source wave fields reversely propagating the last two moments.

4. The method of claim 3, wherein propagating single shot data back into the subsurface from a maximum receive time comprises:

regularizing single shot data;

generating a real speed model in the aperture range according to the aperture range;

and according to the regularized single shot data and the real speed model, reversely propagating the single shot data received from the earth surface back to the underground from the maximum receiving time.

5. The method as claimed in claim 4, wherein the preliminary medical imaging is performed according to the reverse time shift imaging condition.

6. A reverse time migration imaging apparatus for medical images, comprising:

the parameter acquisition module is used for acquiring model parameters and determining a time continuation step length and an offset aperture range according to the model parameters;

the model generation module is used for adding a random speed boundary in the aperture range and generating a random speed boundary model in the aperture range;

the wave field acquisition module is used for determining the maximum receiving time, acquiring forward-playing seismic source wave fields of the last two moments of backward propagation according to the maximum receiving time, and determining the maximum receiving time through the time continuation step length;

the back propagation module is used for back propagating the single shot data to the underground from the maximum receiving time;

a preliminary imaging module for performing a preliminary medical imaging;

the storage module is used for storing a preliminary medical imaging result;

the denoising module is used for removing imaging noise of a primary medical imaging result;

and the imaging determining module is used for recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

7. The apparatus of claim 6, wherein the model parameters comprise a migration parameter and single shot data.

8. The apparatus as claimed in claim 7, wherein the wave field obtaining module comprises:

a wavelet generating unit for generating a source wavelet;

the forward modeling unit is used for forward modeling the seismic source wave field to the maximum receiving time by using a finite difference method according to the seismic source wavelet and the random velocity boundary model;

and the acquisition unit is used for acquiring forward seismic source wave fields of the last two moments of backward propagation.

9. The apparatus as claimed in claim 8, wherein the back propagation module comprises:

the regularization unit is used for regularizing the single shot data;

the model generating unit is used for generating a real speed model in the aperture range according to the aperture range;

and the back propagation unit is used for back propagating the single shot data received from the earth surface back to the underground from the maximum receiving time according to the regularized single shot data and the real speed model.

10. The apparatus of claim 9, wherein the preliminary medical imaging is performed according to a reverse time shift imaging condition.

Technical Field

The invention relates to the field of medical imaging, in particular to a medical image reverse time migration imaging method and device.

Background

Ultrasound (US) medicine is a combined discipline of acoustics, medicine, optics and electronics, and is ultrasonic medicine, which is an application of an acoustic technology for researching frequencies higher than audible sound frequencies in the medical field. Including ultrasonic diagnosis, ultrasonic treatment and biomedical ultrasonic engineering, so that the ultrasonic medicine has the characteristics of combining medical science, theory and engineering, has wide related contents and has high value in preventing, diagnosing and treating diseases.

Ultrasonic imaging is to scan human body with ultrasonic sound beam, receive and process reflected signal to obtain image of internal organs. There are a number of commonly used ultrasound instruments: the type a (amplitude modulation type) indicates the strength of the reflected signal with the amplitude, and a "echo diagram" is shown. The M-mode (spot scanning mode) represents the spatial position from shallow to deep in the vertical direction and time in the horizontal direction, and is shown as a graph of the movement of the spot at different times. The two types are displayed in one dimension, and the application range is limited. Type B (brightness modulation type), namely ultrasonic section imager, is called B-ultrasonic for short. The light spots with different brightness are used for representing the strength of the received signal, when the probe moves along the horizontal position, the light spots on the display screen also move synchronously along the horizontal direction, and the light spot tracks are connected into a sectional view scanned by the ultrasonic sound beams, so that two-dimensional imaging is realized. The D-mode is made according to the ultrasonic Doppler principle, and the C-mode is a scanning mode similar to a television and displays a transverse section acoustic image perpendicular to an acoustic beam. In recent years, ultrasonic imaging techniques such as gray scale display and color display, real-time imaging, ultrasonic holography, transmission ultrasonic imaging, ultrasound parallel tomography, three-dimensional imaging, ultrasonic imaging in body cavities, and the like have been developed. The ultrasonic imaging method is commonly used for judging the position, size and shape of an organ, determining the range and physical properties of a focus, providing an anatomical map of glandular tissues and identifying the normality and abnormality of a fetus, and is widely applied to ophthalmology, obstetrics and gynecology, cardiovascular systems, digestive systems and urinary systems.

However, in the conventional ultrasonic medical imaging, only the primary reflected wave is simply utilized, the phenomena of more complex wave fields, such as reflection, diffraction, multiple waves and the like, are not fully utilized, and the complex information contained in the wave fields is not extracted. Such as velocity, density, etc. of the wave propagation medium, complex structures (e.g., the skull) cannot be accurately imaged.

Disclosure of Invention

In order to solve the technical problems, the method and the device for the reverse time migration imaging of the medical images can realize the medical imaging with low cost and high resolution, and can realize the imaging precision (the precision can reach 0.5 mm) of the nuclear magnetic resonance with the hardware cost of the ultrasonic imaging, even higher precision.

In order to achieve the above purposes, the technical scheme adopted by the invention is as follows:

the invention provides a medical image reverse time migration imaging method, which comprises the following steps:

obtaining model parameters, and determining a time continuation step length and an offset aperture range according to the model parameters;

adding a random speed boundary in the aperture range to generate a random speed boundary model in the aperture range;

determining the maximum receiving time, acquiring forward source wave fields of the last two moments of backward propagation according to the maximum receiving time, and determining the maximum receiving time through the time continuation step length;

from the maximum receive time, back-propagating the single shot data to the subsurface;

performing preliminary medical imaging;

storing preliminary medical imaging results;

removing imaging noise of a preliminary medical imaging result;

and recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

Optionally, the model parameters include an offset parameter and single shot data.

Optionally, determining a maximum receiving time, and obtaining forward source wavefields backward-propagated at the last two moments according to the maximum receiving time, specifically including the following steps:

generating a seismic source wavelet;

forward modeling a seismic source wave field to the maximum receiving time by using a finite difference method according to the seismic source wavelet and the random velocity boundary model;

and acquiring forward seismic source wave fields reversely propagating the last two moments.

Optionally, the back propagation of the single shot data to the underground from the maximum reception time specifically includes the following steps:

regularizing single shot data;

generating a real speed model in the aperture range according to the aperture range;

and according to the regularized single shot data and the real speed model, reversely propagating the single shot data received from the earth surface back to the underground from the maximum receiving time.

Optionally, preliminary medical imaging is performed based on the reverse time-shifted imaging conditions.

Further, the present invention provides a medical image reverse time migration imaging apparatus, comprising:

the parameter acquisition module is used for acquiring model parameters and determining a time continuation step length and an offset aperture range according to the model parameters;

the model generation module is used for adding a random speed boundary in the aperture range and generating a random speed boundary model in the aperture range;

the wave field acquisition module is used for determining the maximum receiving time, acquiring forward-playing seismic source wave fields of the last two moments of backward propagation according to the maximum receiving time, and determining the maximum receiving time through the time continuation step length;

the back propagation module is used for back propagating the single shot data to the underground from the maximum receiving time;

a preliminary imaging module for performing a preliminary medical imaging;

the storage module is used for storing a preliminary medical imaging result;

the denoising module is used for removing imaging noise of a primary medical imaging result;

and the imaging determining module is used for recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

Optionally, the model parameters include an offset parameter and single shot data.

Optionally, the wave field obtaining module specifically includes:

a wavelet generating unit for generating a source wavelet;

the forward modeling unit is used for forward modeling the seismic source wave field to the maximum receiving time by using a finite difference method according to the seismic source wavelet and the random velocity boundary model;

and the acquisition unit is used for acquiring forward seismic source wave fields of the last two moments of backward propagation.

Optionally, the back propagation module specifically includes:

the regularization unit is used for regularizing the single shot data;

the model generating unit is used for generating a real speed model in the aperture range according to the aperture range;

and the back propagation unit is used for back propagating the single shot data received from the earth surface back to the underground from the maximum receiving time according to the regularized single shot data and the real speed model.

Optionally, preliminary medical imaging is performed based on the reverse time-shifted imaging conditions.

The invention has the beneficial effects that:

the invention can realize low-cost high-resolution medical imaging, and can realize the imaging precision (the precision can reach 0.5 mm) of nuclear magnetic resonance and even higher precision with the hardware cost of ultrasonic imaging. Compared to CT, there is no radiation. Compared with the traditional ultrasonic imaging, the method can utilize complex wave fields of reflection, refraction, diffraction, multiple times and the like, has higher imaging precision, and can still provide high-precision quasi-imaging results under the condition that the traditional ultrasonic wave of special parts (such as the skull) cannot be accurately imaged. After the invention is used, the acquisition efficiency is high, one-time acquisition generally needs 2-5 seconds and is not very sensitive to human body movement.

Drawings

FIG. 1 is a flow chart of a method for reverse time migration imaging of medical images according to the present invention;

FIG. 2 is a schematic view of a detailed flow of S300) in the present invention;

FIG. 3 is a schematic flow chart of S400) in the present invention;

FIG. 4 is a block diagram of the medical image reverse time shift imaging device according to the present invention;

FIG. 5 is a block diagram of a specific structure of a wavefield acquisition module in the present invention;

fig. 6 is a specific structural block diagram of the back propagation module in the present invention.

Detailed Description

The following description is presented to disclose the invention so as to enable any person skilled in the art to practice the invention. The preferred embodiments in the following description are given by way of example only, and other obvious variations will occur to those skilled in the art.

The reverse time migration imaging is to utilize the speed model obtained by inversion to perform reverse time migration imaging on the acquired data to obtain the final image.

Reverse time migration (reverse time migration) was first proposed by hemon in 1978 and applied to post-stack migration of longitudinal wave data by Baysal, whitecore, McMechan, and the like, to achieve good results.

The reverse time migration imaging method is to realize migration by combining imaging conditions through the forward propagation of a manually given seismic source wavelet and the backward propagation of received seismic data in a time domain through a two-way wave equation (Claerbout, 1971). Therefore, on one hand, because the reverse time migration needs to calculate the wave field of each time in the forward propagation process of the seismic source and the backward propagation process of the seismic data at the same time, the propagation process of one direction needs to be stored in practical application because the extension directions of the respective times are different, which is why a very large amount of additional storage space is needed. Today, seismic data acquisition has been developed in three or even four dimensions, and the data volume is very large, and the reverse time migration is faced with the known difficulties in the industry when being pushed to practical application. On the other hand, compared with the wave field continuation of the one-way wave equation, the reverse time migration uses the two-way wave fluctuation equation to carry out the wave field continuation, and the reverse time migration becomes the most accurate imaging algorithm because the separation processing of the upper traveling wave and the lower traveling wave is avoided, is not limited by the inclination angle, and can realize the imaging of the rotating wave and the multiple waves. Based on this, geophysicists never forgo pursuit of reverse time migration. Research has also focused on improving the computational efficiency of reverse time migration and reducing storage I/O problems in migration. The establishment of imaging conditions is one of the keys of a seismic migration imaging algorithm, and directly influences the imaging effect and the calculation cost. Imaging conditions were proposed by claerbb, who proposed two imaging conditions simultaneously: the two imaging conditions are called multiplication imaging condition and division imaging condition respectively.

High-order finite difference approximation of the wave equation:

the three-dimensional acoustic wave equation is expressed as follows:

wherein t represents a time vector coordinate, x, y, z represents a three-dimensional space vector coordinate, P represents a displacement function, and V represents a velocity function. Respectively representing approximation by finite difference of the temporal second-order centerUsing spatial high-order-centric finite difference approximationThe following can be obtained:

wherein the content of the first and second substances,

and then approximating by using the N-order central difference:

wherein:

the distances f of the differential grids are respectively expressed by Deltax, Delay and Deltaz_maxRepresenting the maximum, V, of the wavelet frequency_minRepresenting the minimum of the velocity model, the dispersion condition in differential format can be expressed as:

where n denotes that a wavelength is sampled with n samples and h denotes the maximum grid spacing.

Then for this algorithm the stability condition can be expressed as:

the expression of the convolution imaging condition (division imaging condition) can be expressed as:

wherein x, y and z represent coordinates of a three-dimensional space vector, U (x, z and t) is an uplink wave field, D (x, z and t) is a downlink wave field, dt is a continuation time step, I (x and z) represents an imaging result, and tmax represents maximum travel time. The physical meaning is that the reflection coefficient can be determined by dividing the reflected wave energy by the incident wave energy, but this imaging condition is a stability problem in practical applications (possibly by zero).

Therefore, a correlation imaging condition (multiplication imaging condition) is generally employed, and the expression thereof can be expressed as:

extrapolation observation wave field P in double-pass wave reverse time migration_R(x, z, t) instead of the upgoing wave, to extrapolate the wavefield P at the seismic source_S(x, z, t) replaces the down wave, so equation (6) becomes:

integrand P in equation (7)_R(x,z,t)P_S(x, z, t) represents that one imaging operation is carried out on the whole wave field at the time t, and the integral shows that the image in the image space I (x, z) is the superposition of the images formed at each time step. Therefore, the cross-correlation imaging condition makes full use of the imaging information, and effectively suppresses imaging noise while enhancing the imaging signal. The imaging steps of the reverse time shift cross-correlation imaging condition are:

firstly, a wave equation forward calculation is completed, and wave field information of each time step is saved. Then, the wavefield is recorded by reverse time extrapolation, storing wavefield information for each time step. And finally, respectively reading the stored seismic source wave field and the recorded wave field at the same time for imaging operation. I.e. P_R(x,z,t)P_S(x, z, t) and then accumulated into the imaging volume.

The conditions of reverse time migration imaging require the use of the source wavefield (through excitation and forward simulation wave propagation) and the recording wavefield (through reverse time backward propagation back to the underground) at the same time, because one is the forward propagating wavefield and the other is the backward propagating wavefield, if two wavefields at the same time are to be obtained simultaneously, the propagation process of one wavefield, i.e. the wavefield distribution at each time, must be stored, which consumes very large storage resources, and this requirement is difficult to satisfy in actual operation. There are some current solutions, mainly including:

1) strategy: the most conceivable method is to simultaneously record from T the seismic source_n(indicating the time of n × dt) continuation to T₀So that no additional storage space is required and a minimum computation O (2N) can be reached, however T_nThe seismic source wavefield at the time is not known data and therefore cannot be directly implemented;

2) strategy: storage T₀To T_nThen back-propagating the recorded wavefield, and when propagating at a certain time instant, reading the excitation point wavefield at that time instant. The method can also meet the requirement of O (2N) calculated quantity, but needs huge disk space to store the wave field, and even if the disk space can meet the requirement, the generated I/O time is very huge;

3) strategy: only storing a few time-point seismic source wave fields, and when the wave field backward continuation is recorded, utilizing the stored seismic source wave field interpolation to approximate the wave field at the current time point and then imaging, but the wave field interpolation of the processing mode is inaccurate firstly, and the calculation amount is increased;

4) strategy: firstly, the wave field of the seismic source is extended to T in the positive direction_nAnd T_n-1At two moments, the two wavefields are used as initial conditions, with simultaneous backward continuation of the recorded wavefields, and imaging conditions are applied with the continuation, so that no additional storage space is required. This strategy can only be applied in media where the density is constant and the excitation point boundary condition is a Dirichlet boundary condition. However, this extended assumption of wavefields is propagated in infinite space, and the usual calculation method is to incorporate artificial boundaries, including free edgesBoundary conditions, absorption boundary conditions, exponential decay boundary conditions, perfect match layer boundary conditions, and the like. In addition to the free boundary conditions, the three other boundary conditions destroy the integrity of the wavefield, making the continuation of the wavefield irreversible. To do this, the wavefield energy can only be inserted at the appropriate spatial locations at the appropriate times to supplement the wavefield integrity, thus requiring storage of the wavefield values at each time at artificial boundaries, which will rise rapidly as the computation scale increases, although less than 2) the storage required in the strategy, but is generally not practical.

5) Strategy: first forward continuation of the wavefield, and recording multiple sets of checkpoints (P)_n，P_n-1) As the initial condition of the wave field at the middle moment, then extending the wave field of the wave detection point reversely to a certain moment, using the nearest check point to forward propagate to obtain the wave field of the source at the moment, then applying the imaging condition, when applying the PML (perfect match) boundary condition, the wave field needs extra storage space and needs a great deal of repeated calculation process, the maximum repeated calculation may reach the order of o (mn), where m is the number of check points. Griewank proposes an optimized check point selection method by using the thought, so that the number of times of repeated calculation can be reduced to O (Nlog)^m) Despite this, the contradiction between time consumption and storage of reverse time migration has not been completely solved, but this method is currently the most economical calculation strategy.

In the 5 solutions of reverse time migration, it can be seen that the computation magnitude and the storage size are contradictory, and if the minimum computation amount is desired, the maximum storage amount is reached, and vice versa. However, in strategy four, no additional storage space is required, and other strategies involve a balance between storage size and computational complexity. In strategy four, artificial boundary is the main influencing factor, however, free boundary condition will introduce much imaging artifact, and other boundary condition will not compound the requirement of the strategy but can only be abandoned.

In the present invention, the above-mentioned strategy 4) is mainly aimed at. In 2009 Robert proposed a stochastic boundary model whose idea was to eliminate the coherence of artificial boundary free boundary condition reflections, making boundary reflections impossible to image. The specific implementation process is as follows: expanding the limited space for a certain distance, then filling random speed in the expanded space to form a random boundary speed model, randomizing the wave front when the wave is transmitted to a random speed region, converting the wave field into random noise and transmitting the random noise back to a real speed region, and destroying the coherence of boundary reflection to make the boundary reflection incapable of imaging.

The random boundary function is constructed as follows:

wherein x, y, z represent space vector coordinates,is the random velocity function of the boundary point, V (x, y) is the original velocity function of the boundary point, r is the random number, d is the spatial distance of the velocity point from the inner layer boundary.

The excitation point wavefield may be propagated to T using a random velocity boundary model using a processing unit, such as a CPU/GPU, as a migration computation core_nTime of day, then using T_nThe wave field of the moment is used as an initial condition, the excitation point wave field and the wave detection point wave field are transmitted back at the same time, and imaging is carried out by using the imaging condition at the same time. And additional storage space is avoided, and although the excitation point wave field is repeatedly calculated once, the time consumption of the GPU used as a numerical calculation core of the wave field continuation is very economical compared with large-scale disk I/O.

In view of the above, referring to fig. 1, the method for reverse time migration imaging of medical images of the present invention includes the following steps:

s100) obtaining model parameters, and determining a time continuation step length and a migration aperture range according to the model parameters, in this embodiment, the model parameters include a migration parameter and single-shot data, and the continuation step length may be determined according to the migration parameter, and then the migration aperture range may be determined according to the single-shot data, in addition, the single-shot data refers to a shot concept in seismic exploration, where after a wave field generating device, such as an ultrasonic transducer, may be launched once, all data recorded by the wave field generating device is a shot, that is, a single-shot data.

S200) adding a random speed boundary in the aperture range to generate a random speed boundary model in the aperture range.

S300) determining the maximum receiving time, and acquiring forward seismic source wave fields (T) of the last two moments of backward propagation according to the maximum receiving time_nAnd T_n-1The forward source wavefield at two time points, the maximum receiving time is determined by the time continuation step, specifically, referring to fig. 2, the method includes the following steps:

s301) generates a source wavelet.

S302) forward modeling the source wave field to the maximum receiving time (GPU acceleration) by using a finite difference method according to the source wavelet and the random velocity boundary model.

S303) obtain forward source wavefields backward propagating the last two time instants (GPU acceleration).

S400) back-propagating the single shot data from the maximum receive time back into the subsurface, specifically, referring to fig. 3, comprising the steps of:

s401) regularizing the single shot data to obtain regularized single shot data.

S402) generating a real speed model in the aperture range according to the aperture range.

S403), according to the regularized single shot data and the real speed model, reversely propagating the single shot data received from the earth surface back to the underground again from the maximum receiving time (GPU acceleration).

S500) performing preliminary medical imaging according to the reverse time migration imaging condition (GPU acceleration).

S600) storing the preliminary medical imaging result.

S700) removing imaging noise of the preliminary medical imaging result.

S800) recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

Further, referring to fig. 4, the medical image reverse time migration imaging apparatus provided by the present invention includes:

the parameter obtaining module is configured to obtain a model parameter, and determine a time continuation step length and a migration aperture range according to the model parameter, in this embodiment, the model parameter includes a migration parameter and single shot data, and the continuation step length may be determined according to the migration parameter, and then the migration aperture range may be determined according to the single shot data, in addition, the single shot data refers to a shot concept in seismic exploration, where after a primary wave field generating device such as an ultrasonic transducer is launched, all data recorded by a wave field receiving generating device is a shot, that is, a single shot data.

And the model generation module is used for adding a random speed boundary in the aperture range and generating a random speed boundary model in the aperture range.

A wave field acquisition module for determining maximum receiving time and acquiring forward seismic source wave field (T) of last two backward propagation moments according to the maximum receiving time_nAnd T_n-1The forward source wavefield at two time points, the maximum receiving time is determined by the time continuation step, specifically, referring to fig. 5, the method includes the following steps:

and the wavelet generating unit is used for generating the source wavelet.

And the forward modeling unit is used for forward modeling the source wave field to the maximum receiving time (GPU acceleration) by using a finite difference method according to the source wavelet and the random velocity boundary model.

And the acquisition unit is used for acquiring forward seismic source wave fields (GPU acceleration) of the last two moments of backward propagation.

A back propagation module for back propagating the single shot data from the maximum receive time back into the subsurface, specifically, referring to fig. 6, comprising the steps of:

and the regularization unit is used for regularizing the single shot data to obtain the regularized single shot data.

And the model generating unit is used for generating a real speed model in the aperture range according to the aperture range.

And the back propagation unit is used for back propagating the single shot data received from the earth surface back to the underground (GPU acceleration) from the maximum receiving time according to the regularized single shot data and the real speed model.

And the preliminary imaging module is used for performing preliminary medical imaging according to the reverse time migration imaging condition (GPU acceleration).

And the storage module is used for storing the preliminary medical imaging result.

And the denoising module is used for removing imaging noise of the preliminary medical imaging result.

And the imaging determining module is used for recovering the spatial coordinate position information of the imaging result to obtain a final medical imaging result.

The invention has the beneficial effects that:

the invention can realize low-cost high-resolution medical imaging, and can realize the imaging precision (the precision can reach 0.5 mm) of nuclear magnetic resonance and even higher precision with the hardware cost of ultrasonic imaging. Compared to CT, there is no radiation. Compared with the traditional ultrasonic imaging, the method can utilize complex wave fields of reflection, refraction, diffraction, multiple times and the like, has higher imaging precision, and can still provide high-precision quasi-imaging results under the condition that the traditional ultrasonic wave of special parts (such as the skull) cannot be accurately imaged. After the invention is used, the acquisition efficiency is high, one-time acquisition generally needs 2-5 seconds and is not very sensitive to human body movement.

The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are merely illustrative of the principles of the invention, but that various changes and modifications may be made without departing from the spirit and scope of the invention, which fall within the scope of the invention as claimed.