CN113633314A - Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing - Google Patents

Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing Download PDF

Info

Publication number
CN113633314A
CN113633314A CN202111041625.9A CN202111041625A CN113633314A CN 113633314 A CN113633314 A CN 113633314A CN 202111041625 A CN202111041625 A CN 202111041625A CN 113633314 A CN113633314 A CN 113633314A
Authority
CN
China
Prior art keywords
ultrasonic
gpu
signal
composite image
complex
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.)
Pending
Application number
CN202111041625.9A
Other languages
Chinese (zh)
Inventor
李振华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Xinhuan Technology Co ltd
Original Assignee
Individual
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN202111041625.9A priority Critical patent/CN113633314A/en
Publication of CN113633314A publication Critical patent/CN113633314A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/52—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/5215—Devices using data or image processing specially adapted for diagnosis using ultrasonic, sonic or infrasonic waves involving processing of medical diagnostic data
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
    • A61B8/4444—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device related to the probe
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/44—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device
    • A61B8/4483—Constructional features of the ultrasonic, sonic or infrasonic diagnostic device characterised by features of the ultrasound transducer
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/48—Diagnostic techniques
    • A61B8/483—Diagnostic techniques involving the acquisition of a 3D volume of data

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Medical Informatics (AREA)
  • Physics & Mathematics (AREA)
  • Surgery (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Gynecology & Obstetrics (AREA)
  • Ultra Sonic Daignosis Equipment (AREA)

Abstract

The invention discloses an image synthesis method for accelerating ultrasonic multi-plane wave imaging by using GPU parallel computation, which comprises the following steps: s1, setting different emission delays of all transducer units of the ultrasonic probe, and sequentially emitting ultrasonic plane waves with different emission angles to scan an imaging space; s2, reflecting the transmitted ultrasonic signals back to the ultrasonic probe, receiving the ultrasonic echo signals by each transducer unit, and converting the ultrasonic echo signals into Radio Frequency (RF) signals after a series of filtering, sampling, gain and other processing in the analog front end; s3, generating a composite image by using the radio frequency signal through the following post-processing steps; vectorizing the radio frequency signals collected by all channels by using a CPU; performing Hilbert transformation accelerated by GPU parallel computing on the radio frequency signals by using a GPU, and realizing composite image synthesis by using a DAS algorithm accelerated by GPU parallel computing; and S4, displaying the synthesized composite image on a display device. The ultrasonic ultra-fast imaging post-processing algorithm has good parallel computing applicability, and can be used for image reconstruction by using high-performance computing platforms such as a GPU (graphics processing unit) and the like so as to improve the computing efficiency of image reconstruction.

Description

Ultrasonic multi-plane wave composite image synthesis method and system based on GPU parallel computing acceleration
Technical Field
The invention relates to the technical field of ultrasonic imaging, in particular to an ultrasonic multi-plane wave composite image synthesis method based on GPU parallel computing acceleration.
Background
In medical ultrasonic imaging, by sequentially emitting and receiving a plurality of unfocused spatial plane waves at different angles, a focus and a tissue structure to be imaged are scanned, and ultra-fast ultrasonic imaging (frame rate >1000fps) can be clinically realized, and the technology is currently often applied to cardiovascular imaging such as cardiac structure, valve structure, blood flow imaging and the like.
The ultrasound imaging commonly used in clinic is two-dimensional imaging (i.e. area imaging) by using an ultrasound probe such as a linear array, a circular array, etc., or three-dimensional imaging (i.e. volume imaging) by using an ultrasound probe such as a two-dimensional area array, a 1.5D probe, etc. There are generally two ways to scan and image the imaging space using these probes, namely dynamic sub-aperture synthesis and multi-plane wave complex imaging (i.e. ultra-fast ultrasound imaging), the dynamic sub-aperture synthesis method is slow in physical imaging and low in data throughput, and the multi-plane wave complex physical imaging is fast but large in data volume and therefore slow in computation.
The ultra-fast ultrasonic imaging scans an imaging space by sequentially emitting N non-aggregated plane waves with different emission angles. When the ultrasonic wave is transmitted at a certain specific transmitting angle, the time Delay of different transducer units is different, and a low-resolution image can be obtained through a DAS algorithm (Delay and Sum); and carrying out composite addition on the obtained N low-resolution images to obtain the high-resolution ultrasonic B-mode image. The method has high scanning and receiving speed and great scientific research and clinical value, but has great data volume, slow calculation, high-performance calculation for accelerating image reconstruction and great clinical significance.
Ultra-fast imaging generally uses all transducer units to perform ultrasonic signal excitation and ultrasonic signal acquisition, the acquired data volume is huge, the existing ultrasonic ultra-fast imaging post-processing method is generally based on an FPGA (Field Programmable Gate Array) or a CPU (central processing unit), and the data throughput and the computing capacity of the FPGA are limited. With the development of ultrasonic imaging technology and the further increase of the number of transducer units, the common FPGA or CPU has been difficult to meet the calculation requirements of real-time image reconstruction.
Disclosure of Invention
In order to improve the image reconstruction speed and the calculation efficiency of the ultrasonic ultra-fast imaging of the multi-plane wave compounding, the invention aims to provide an ultrasonic multi-plane wave compounding image synthesis method which uses parallel calculation to accelerate the multi-plane wave compounding and provides a GPU (graphic processing unit) parallel calculation to accelerate.
The invention is realized by the following technical scheme.
The technical scheme of the invention is an ultrasonic multi-plane wave composite image synthesis method for accelerating GPU parallel computation, which comprises the following steps:
s1, sequentially adopting different channels to transmit time delay to different transducer units of the ultrasonic probe, and transmitting ultrasonic signals of non-aggregated ultrasonic plane waves to an imaging space region at different transmitting angles;
s2, transmitting the ultrasonic wave in the tissue and reflecting the ultrasonic wave back to the ultrasonic probe, and obtaining a radio frequency signal through a series of analog-to-digital conversion, filtering and gain;
s3, generating a composite image by using the radio frequency signal through the following post-processing steps;
vectorizing the radio frequency signals obtained by all the transducer channels by using a CPU (central processing unit) and transmitting the radio frequency signals to the CPU;
performing Hilbert transform accelerated by GPU parallel computation on the vector-quantized radio-frequency signals by using a GPU, and realizing composite image synthesis by using a DAS algorithm accelerated by GPU parallel computation;
s4, the composite image after the synthesis is transferred back to the CPU and displayed on the display device.
Compared with the prior art, the computation efficiency of high-performance parallel computation based on the GPU is shortened by hundreds or even thousands of times compared with the running time of the CPU, the ultrasonic ultra-fast imaging post-processing algorithm has good parallel computation applicability, and high-performance computing platforms such as the GPU can be used for image reconstruction to improve the computation efficiency.
In one embodiment of this solution, in step S1, the transducer unit emits ultrasonic waves in a plane wave manner, the emitted ultrasonic waves being unfocused plane ultrasonic waves having different emission angles.
In one embodiment of this embodiment, the "processing the ultrasonic wave transmission signal and the received ultrasonic wave reception signal using the CPU to obtain the radio frequency signal" in step S2 includes:
and respectively filtering, sampling, time gain compensation and analog-to-digital conversion processing echo signals of all channels acquired by the ultrasonic simulation front-end system to obtain radio frequency signals, and transmitting the radio frequency signals to the CPU.
In an embodiment of this embodiment, in step 3, the hilbert transform for performing GPU parallel computation acceleration on the rf signal by using the GPU includes:
performing discrete Fourier transform on the opposite-quantization radio frequency signal to obtain a complex signal F;
carrying out different phase shifts on different sampling points in the complex signal F to obtain a real signal P;
performing inverse discrete Fourier transform on the phase-shifted real signal to obtain a complex signal I;
further, when different phase shifts are performed on different sampling points in the reset signal F, and when the number NS of the sampling points is an even number, the formula (1) is followed; when odd, then according to equation (2)), where i is in imaginary units:
Figure BDA0003249513670000031
Figure BDA0003249513670000032
and performing GPU-accelerated inverse discrete Fourier transform on the phase-shifted real signal P to obtain a complex signal I, wherein the complex signal obtained by Hilbert transform is S ═ R + I × (I), and abs represents a modulus of the complex number.
In an embodiment of this embodiment, in step 3, the DAS algorithm with GPU parallel computation acceleration performs composite image synthesis including:
carrying out interpolation and time delay on a complex signal S obtained after Hilbert transformation according to the total time delay of the voxel point in each channel, and adding interpolation results on all channels to obtain a corresponding low-resolution complex image; and carrying out complex addition on the plurality of low-resolution complex images to obtain high-resolution complex images, wherein the complex modulus of the complex images is the finally synthesized B-mode image.
Further, the interpolation and the delay addition are performed on the complex signal obtained after the hilbert transform to obtain a corresponding complex image with low resolution, and the method includes:
calculating emission delay according to a plane wave emission angle, calculating receiving delay corresponding to the channel of the ultrasonic transducer according to the spatial relative positions of the voxel points and the ultrasonic transducers of the channels, wherein the total delay is the sum of the emission delay and the receiving delay, and a total delay value is obtained by each channel corresponding to each voxel point;
according to the total delay on each ultrasonic receiving channel corresponding to each voxel point, performing interpolation calculation on the complex signal S corresponding to the channel, and adding the complex values acquired by the voxel point on all channels to obtain a low-resolution complex image of the voxel point;
and respectively calculating the plurality of emission angles, wherein each time the emission angles are different, repeating the steps to obtain a plurality of low-resolution complex images.
Drawings
The foregoing and other objects, features and advantages of the application will be apparent from the following more particular descriptions of exemplary embodiments of the application, as illustrated in the accompanying drawings wherein like reference numbers generally represent like parts throughout the exemplary embodiments of the application.
Fig. 1 is a flow chart of GPU parallel computation accelerated ultrasound image synthesis in an embodiment.
Detailed Description
Preferred embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While the preferred embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The embodiment is an ultrasonic multi-plane wave composite image synthesis method for accelerating GPU parallel computing, which comprises the following steps:
s1, sequentially transmitting non-aggregated plane waves with different transmitting angles to an imaging space region by controlling transmitting time delay of each transducer of the ultrasonic probe;
s2, transmitting and reflecting the ultrasonic signals in the tissues to the ultrasonic probe, and performing a series of operations such as filtering, gain, analog-to-digital conversion and the like on the received echo signals in the analog front end to obtain radio frequency signals;
s3, generating a composite image by using the radio frequency signal through the following post-processing steps;
vectorizing the radio-frequency signals received by each transducer channel by using a CPU (central processing unit);
performing Hilbert transform accelerated by GPU parallel computation on the vector-quantized radio-frequency signals by using a GPU, and realizing composite image synthesis by using a DAS algorithm accelerated by GPU parallel computation;
and S4, displaying the synthesized composite image on a display device.
Referring to fig. 1, the detailed implementation process of the present embodiment is as follows:
assuming that there are N ultrasound emission angles and NC ultrasound transducer channels, taking a three-dimensional imaging mode of a two-dimensional area array or a 1.5D ultrasound probe as an example, the imaging space region has NX × NY × NZ voxel points in total, and in addition, if a two-dimensional imaging mode such as a linear array, a circular array, etc., the imaging space region has NX × NZ voxel points.
And transmitting at the kth ultrasonic transmitting angle, controlling the transmitting time delay of each transducer, respectively performing a series of filtering, time gain compensation and analog-to-digital conversion processing operations on the ultrasonic signals of each channel acquired by the ultrasonic analog front-end system to obtain radio frequency signals, recording the radio frequency signals as RF signals, and transmitting the RF signals into a CPU (central processing unit) of the transducer module through data interfaces such as PCI-e (peripheral component interconnect-express) and the like.
Setting the total sampling number of each channel as NS, vectorizing the RF signals collected by NC channels on a CPU to obtain a one-dimensional vector of NS NC, and transmitting the one-dimensional vector to a GPU;
performing rapid Hilbert transform on the vector on a GPU for GPU parallel computation acceleration, wherein the number of parallel threads is NS NC, and the Hilbert transform on the GPU comprises the following three steps:
firstly, performing GPU accelerated discrete Fourier transform on a vector quantized RF signal to obtain a complex signal F;
and secondly, performing different phase shifts on different sampling points in the F (according to a formula (1) if NS is even number, or according to a formula (2) if NS is odd number), wherein i is an imaginary unit:
Figure BDA0003249513670000041
Figure BDA0003249513670000042
and thirdly, performing GPU-accelerated inverse discrete Fourier transform on the phase-shifted real signal P to obtain a complex signal I, wherein the complex signal obtained by Hilbert transform is S ═ R + I × (I), and abs represents the modulus of the complex number.
DAS (delay And Sum) image synthesis with GPU parallel computing acceleration is carried out. In this embodiment, taking ultrasound three-dimensional volume imaging as an example, the imaging space has NX × NY × NZ voxel points in total, and each voxel point is parallelized, so that there are NX × XY × NZ threads in total, and for each thread, the following calculation is performed:
in the first step, the transmit and receive delays of the voxel point are calculated.
Calculating the emission delay according to the emission angle of the current plane wave; calculating receiving delay according to the space relative position of each channel of the voxel point and the ultrasonic transducer, wherein the total delay is the sum of transmitting delay and receiving delay, and a total delay value is obtained by each channel (NC channels in total) corresponding to each voxel point;
and secondly, interpolating.
And performing linear interpolation on the complex signal S corresponding to each channel according to the total delay on each channel corresponding to each voxel point. Each thread adds the values obtained on the NC channels to obtain a low-resolution complex image corresponding to the voxel point;
and thirdly, synthesizing images.
Respectively calculating N different emission angles, taking the current emission angle for each calculation, performing the first step and the second step to obtain N low-resolution complex images, adding all the low-resolution complex images to obtain a high-resolution complex image, and taking the final image as a complex modulus of the complex image.
And transmitting the synthesized final B-mode image back to the CPU from the GPU, and performing user side operation such as image output and display on a display device.
The present embodiment is verified by using a 1024 ultrasonic area array probe, the running time of the present embodiment is shortened by hundreds or even thousands of times compared to that of a CPU, the NVIDIA 3090ti GPU-based test time of ultrasonic image reconstruction using 1024 area arrays and 30 plane waves is about 0.6 seconds, while the time is about 40 minutes on the CPU, and the CPU and the GPU calculate results are consistent. The results are proved to be feasible through algorithm verification and calculation and comparison on different platforms.
Having described embodiments of the present application, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims (8)

1.一种GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，包括有如下步骤：1. an ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing, is characterized in that, comprises the following steps: S1、对超声探头的各个换能器单元，不同通道采用发射延时，从而以不同发射角度发射非聚集超声平面波扫描成像空间；S1. For each transducer unit of the ultrasonic probe, different channels adopt transmission delay, so as to transmit non-aggregated ultrasonic plane waves at different transmission angles to scan the imaging space; S2、超声信号在组织中传播并反射，超声探头接受到回波信号后，在模拟前端中经过处理后获得射频信号；S2. The ultrasonic signal is propagated and reflected in the tissue. After the ultrasonic probe receives the echo signal, the radio frequency signal is obtained after processing in the analog front end; S3、通过如下后处理步骤，使用射频信号产生复合图像；S3, using the radio frequency signal to generate a composite image through the following post-processing steps; 使用CPU对超声探头上各个换能器通道获得的射频信号进行向量化，并传输数据到GPU中；Use the CPU to vectorize the RF signals obtained by each transducer channel on the ultrasound probe, and transmit the data to the GPU; 使用GPU对向量化的射频信号进行GPU并行计算加速的希尔伯特变换，以及GPU并行计算加速的DAS算法，实现复合图像合成；Use GPU to perform Hilbert transform accelerated by GPU parallel computing on the vectorized radio frequency signal, and DAS algorithm accelerated by GPU parallel computing to realize composite image synthesis; S4、将合成后的复合图像传输回CPU，并在显示装置上显示。S4. The synthesized composite image is transmitted back to the CPU, and displayed on the display device. 2.根据权利要求1所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，在步骤S1中，换能器单元通过控制不同换能器通道的发射延时，依次以不同发射角度发射非聚焦超声平面波，对成像空间进行扫描。2. The ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing according to claim 1, is characterized in that, in step S1, the transducer unit controls the transmission delay of different transducer channels successively with different The emission angle emits unfocused ultrasonic plane waves to scan the imaging space. 3.根据权利要求1所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，在步骤S2中，“超声信号在组织中传播并反射，超声探头接受到回波信号后，处理后获得射频信号；”包括：3. the ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing according to claim 1, is characterized in that, in step S2, " ultrasonic signal is propagated and reflected in tissue, after ultrasonic probe receives echo signal, After processing to obtain a radio frequency signal;" includes: 将超声模拟前端系统采集到的各个通道的回波信号分别进行滤波、采样、时间增益补偿和模数转换处理后，获得射频信号，并将射频信号传递到CPU。After the echo signals of each channel collected by the ultrasonic analog front-end system are filtered, sampled, time gain compensation and analog-to-digital conversion processing, the radio frequency signal is obtained, and the radio frequency signal is transmitted to the CPU. 4.根据权利要求3所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，在步骤3中，其中使用GPU对向量化的射频信号进行GPU并行计算加速的希尔伯特变换包括有：4. the ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing according to claim 3, is characterized in that, in step 3, wherein uses GPU to carry out the Hilbert of GPU parallel computing acceleration to the radio frequency signal of vectorization Transformations include: 将所有通道获得的射频信号在CPU上进行向量化，并传递到GPU中；The RF signals obtained by all channels are vectorized on the CPU and passed to the GPU; 对向量化的射频信号进行离散傅里叶变换，获得复数信号F；Perform discrete Fourier transform on the quantized radio frequency signal to obtain a complex signal F; 对复数信号F中不同采样点进行不同相移，得到实信号P；Different phase shifts are performed on different sampling points in the complex signal F to obtain the real signal P; 对相移后的实信号进行离散傅里叶逆变换，得至复数信号I。Perform inverse discrete Fourier transform on the phase-shifted real signal to obtain the complex signal I. 5.根据权利要求4所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，其中对复位信号F中不同采样点进行不同相移时，当采样点的数量NS为偶数，则根据公式(1)；当为奇数时，则根据公式(2))，其中i为虚数单位：5. the ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing according to claim 4, is characterized in that, when wherein different sampling points in reset signal F are carried out different phase shifts, when the quantity NS of sampling points is an even number, Then according to formula (1); when it is odd, according to formula (2)), where i is the imaginary unit:
Figure FDA0003249513660000021
Figure FDA0003249513660000021
Figure FDA0003249513660000022
Figure FDA0003249513660000022
 对相移后的实信号P进行GPU加速的离散傅里叶逆变换，得到复数信号I，希尔伯特变换获得的复数信号为S＝R+i*abs(I)，其中abs表示复数的模。Perform GPU-accelerated inverse discrete Fourier transform on the phase-shifted real signal P to obtain a complex signal I, and the complex signal obtained by Hilbert transform is S=R+i*abs(I), where abs represents the complex number mold. 6.根据权利要求4或5所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，在步骤3中，其中GPU并行计算加速的DAS算法实现复合图像合成包括有：6. the ultrasonic multi-plane wave composite image synthesis method of GPU parallel computing acceleration according to claim 4 or 5, is characterized in that, in step 3, wherein the DAS algorithm of GPU parallel computing acceleration realizes composite image synthesis and comprises: 对希尔伯特变换后获得的复数信号S进行插值和延时相加，得到一个对应的低分辨率的复数图像；对多个低分辨率的复数图像进行复数相加，获得高分辨率的复数图像，复数图像的复数模即为最终合成的B模式图像。Perform interpolation and delay addition on the complex signal S obtained after the Hilbert transform to obtain a corresponding low-resolution complex image; perform complex addition on multiple low-resolution complex images to obtain a high-resolution complex image. For complex images, the complex modulus of the complex images is the final synthesized B-mode image. 7.根据权利要求6所述的GPU并行计算加速的超声多平面波复合图像合成方法，其特征在于，其中对希尔伯特变换后获得的复数信号进行插值和延时相加，得到一个对应的低分辨率的复数图像，包括有：7. The ultrasonic multi-plane wave composite image synthesis method accelerated by GPU parallel computing according to claim 6, is characterized in that, wherein the complex signal obtained after the Hilbert transform is interpolated and added by time delay, obtaining a corresponding Low-resolution complex images, including: 根据平面波发射角度计算发射延时，根据体素点和各通道超声换能器的空间相对位置计算接收延时，总延时为发射延时和接收延时之和，每个体素点对应每个通道都会得到一个总延时值；The transmission delay is calculated according to the plane wave transmission angle, and the reception delay is calculated according to the spatial relative position of the voxel point and the ultrasonic transducer of each channel. The total delay is the sum of the transmission delay and the reception delay. Each voxel point corresponds to each Each channel will get a total delay value; 根据各个体素点对应的各个超声接收通道上的总延时，在该通道对应的复数信号S上进行插值计算，并对该体素点在所有通道上获取的复数值相加，即可得到低分辨率的复数图像；According to the total delay on each ultrasonic receiving channel corresponding to each voxel point, perform interpolation calculation on the complex signal S corresponding to this channel, and add the complex values obtained by this voxel point on all channels to obtain low-resolution complex images; 分别对不同发射角度进行计算，每次计算根据发射角度，重复前述步骤，得到多个低分辨率的复数图像。Calculations are performed for different emission angles respectively, and the foregoing steps are repeated according to the emission angles for each calculation to obtain multiple low-resolution complex images. 8.一种GPU并行计算加速的超声多平面波复合图像合成系统，其特征在于，包括有：8. an ultrasonic multi-plane wave composite image synthesis system accelerated by GPU parallel computing, is characterized in that, comprises: 超声模块，对超声探头的不同换能器单元，采用不同通道发射延时，以不同发射角度发射非聚集超声平面波的超声信号至成像空间区域中；S2、超声信号在组织中传播并反射，超声探头接受到回波信号后，处理后获得射频信号；The ultrasonic module uses different channel transmission delays for different transducer units of the ultrasonic probe, and transmits ultrasonic signals of non-aggregated ultrasonic plane waves into the imaging space area at different emission angles; S2, the ultrasonic signals propagate and reflect in the tissue, and the ultrasonic waves After the probe receives the echo signal, the radio frequency signal is obtained after processing; 处理器模块，包括有CPU，使用CPU对每个射频信号进行向量化；The processor module, including a CPU, uses the CPU to vectorize each RF signal; 处理器模块，包括有GPU，使用GPU对射频信号进行GPU并行计算加速的希尔伯特变换，以及GPU并行计算加速的DAS算法实现复合图像合成；The processor module includes a GPU, which uses the GPU to perform the Hilbert transform accelerated by GPU parallel computing on the radio frequency signal, and the DAS algorithm accelerated by the GPU parallel computing to realize composite image synthesis; 显示模块，将合成后的复合图像在显示装置上显示。The display module displays the synthesized composite image on the display device.
CN202111041625.9A 2021-09-07 2021-09-07 Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing Pending CN113633314A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111041625.9A CN113633314A (en) 2021-09-07 2021-09-07 Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111041625.9A CN113633314A (en) 2021-09-07 2021-09-07 Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing

Publications (1)

Publication Number Publication Date
CN113633314A true CN113633314A (en) 2021-11-12

Family

ID=78425210

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111041625.9A Pending CN113633314A (en) 2021-09-07 2021-09-07 Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing

Country Status (1)

Country Link
CN (1) CN113633314A (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115990039A (en) * 2022-12-13 2023-04-21 深圳心寰科技有限公司 Ultrasonic imaging method, device, equipment and medium for multi-beam focusing emission
CN117710184A (en) * 2024-01-04 2024-03-15 浙江大学 Ultrasonic full-focusing wave number domain imaging method, system and device based on FPGA and GPU
WO2025073121A1 (en) * 2023-10-03 2025-04-10 Hong Kong Applied Science And Technology Research Institute Co., Ltd. Image processing method for multi-frequency ultrasound probe device
WO2025076977A1 (en) * 2023-10-11 2025-04-17 深圳心寰科技有限公司 Four-dimensional ultrasound imaging system and method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN87102652A (en) * 1986-02-21 1987-09-30 老代尔夫特光学工业有限公司 The ultrasonic listening Apparatus and method for
CN104546003A (en) * 2015-01-27 2015-04-29 哈尔滨工业大学 VFR (variable frame rate) color ultra-high speed ultrasonic imaging method based on plane wave transmission
CN104688273A (en) * 2015-03-16 2015-06-10 哈尔滨工业大学 Ultra high speed ultrasonic imaging device and method based on central processing unit (CPU) + graphic processing unit (GPU) isomeric framework
US20150327840A1 (en) * 2014-05-19 2015-11-19 Kabushiki Kaisha Toshiba Ultrasonic diagnostic device and correction method
CN109788941A (en) * 2016-09-28 2019-05-21 柯惠有限合伙公司 The system and method for CPU and GPU processing parallelization for supersonic imaging apparatus
CN112754529A (en) * 2021-01-08 2021-05-07 大连东软教育科技集团有限公司 Ultrasonic plane wave imaging method and system based on frequency domain migration and storage medium

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN87102652A (en) * 1986-02-21 1987-09-30 老代尔夫特光学工业有限公司 The ultrasonic listening Apparatus and method for
US20150327840A1 (en) * 2014-05-19 2015-11-19 Kabushiki Kaisha Toshiba Ultrasonic diagnostic device and correction method
CN104546003A (en) * 2015-01-27 2015-04-29 哈尔滨工业大学 VFR (variable frame rate) color ultra-high speed ultrasonic imaging method based on plane wave transmission
CN104688273A (en) * 2015-03-16 2015-06-10 哈尔滨工业大学 Ultra high speed ultrasonic imaging device and method based on central processing unit (CPU) + graphic processing unit (GPU) isomeric framework
CN109788941A (en) * 2016-09-28 2019-05-21 柯惠有限合伙公司 The system and method for CPU and GPU processing parallelization for supersonic imaging apparatus
CN112754529A (en) * 2021-01-08 2021-05-07 大连东软教育科技集团有限公司 Ultrasonic plane wave imaging method and system based on frequency domain migration and storage medium

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
殷福亮，宋爱军: "《数字信号处理C语言程序集》", 31 July 1997, 辽宁科学技术出版社, pages: 140 - 144 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115990039A (en) * 2022-12-13 2023-04-21 深圳心寰科技有限公司 Ultrasonic imaging method, device, equipment and medium for multi-beam focusing emission
WO2025073121A1 (en) * 2023-10-03 2025-04-10 Hong Kong Applied Science And Technology Research Institute Co., Ltd. Image processing method for multi-frequency ultrasound probe device
WO2025076977A1 (en) * 2023-10-11 2025-04-17 深圳心寰科技有限公司 Four-dimensional ultrasound imaging system and method
CN117710184A (en) * 2024-01-04 2024-03-15 浙江大学 Ultrasonic full-focusing wave number domain imaging method, system and device based on FPGA and GPU

Similar Documents

Publication Publication Date Title
CN113633314A (en) Ultrasonic multi-plane wave composite image synthesis method and system accelerated by GPU parallel computing
US20230000472A1 (en) Ultrasonic imaging compression methods and apparatus
CN104739448B (en) A kind of ultrasonic imaging method and device
US20130172752A1 (en) Ultrasound transducer apparatus and ultrasound imaging system and imaging method
Zhang et al. Extension of Fourier-based techniques for ultrafast imaging in ultrasound with diverging waves
EP3824313A1 (en) Ultrasound imaging by deep learning and associated devices, systems, and methods
EP2610641B1 (en) Ultrasound and system for forming a Doppler ultrasound image
US20180092627A1 (en) Ultrasound signal processing device, ultrasound signal processing method, and ultrasound diagnostic device
US10575823B2 (en) Medical diagnostic apparatus, medical image processing apparatus and medical image processing method
Cohen et al. Sparse convolutional beamforming for 3-D ultrafast ultrasound imaging
CN103948399A (en) Three-dimensional ultrasonic imaging method based on non-diffraction waves in sector scanning mode
CN102188261B (en) Synthetic aperture three-dimensional ultrasonic imaging method based on non-diffraction array waves
JP2003180686A (en) Method and device for measurement of displacement vector and device for measurement of distortion tensor
Angiolini et al. 1024-Channel 3D ultrasound digital beamformer in a single 5W FPGA
Bosisio et al. Real-time chirp-coded imaging with a programmable ultrasound biomicroscope
WO2019202593A1 (en) Sparse convolutional beamforming for ultrasound imaging
Bueno et al. Real-time ultrasound image reconstruction as an inverse problem on a GPU
Ibrahim et al. Apodization scheme for hardware-efficient beamformer
Pospisil et al. 4-D x 3-D ultrasound: real-time scan conversion, filtering, and display of displacement vectors with a motorized curvilinear transducer
Guo et al. Wavenumber Beamforming with Sub-Nyquist Sampling for Focus-Beam Ultrasound Imaging
Han et al. A two-steps implementation of 3D ultrasound imaging in frequency domain with 1D array transducer
Kassem et al. A new digital scan conversion architecture for ultrasonic imaging systems
CN115363627B (en) Array imaging method, device, storage medium and electronic equipment
CN115115731B (en) Software beamforming method, device, ultrasonic imaging device and storage medium
CN120052948A (en) Ultrasonic imaging method and ultrasonic imaging system

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
TA01 Transfer of patent application right
TA01 Transfer of patent application right

Effective date of registration: 20221123

Address after: 518,000 506, 507, Building 2, Bao'an Foreign Trade Industrial Zone, No. 51, Lane 1, Xin'an Third Road, Dalang Community, Bao'an District, Shenzhen, Guangdong

Applicant after: Shenzhen Xinhuan Technology Co.,Ltd.

Address before: 518000 West New City Mansion, Shennan Middle Road, Futian District, Shenzhen, Guangdong 501

Applicant before: Li Zhenhua