Ultrasonic plane wave composite imaging method based on deep learning
A technology of deep learning and composite imaging, applied in neural learning methods, ultrasonic/sonic/infrasonic image/data processing, ultrasonic/sonic/infrasonic diagnosis, etc., can solve problems such as difficult image reconstruction, achieve higher frame rate, lower Difficulty, the effect of reducing complexity
Pending Publication Date: 2021-11-19
HEFEI UNIV OF TECH
View PDF0 Cites 2 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Many existing deep learning methods also learn the weight of channel data. Once the scene changes, it is difficult to reconstruct a better image.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0045] In this embodiment, an ultrasonic plane wave compound imaging method based on deep learning, the specific process is as follows figure 1 shown, including the following steps:
[0046] Step 1: Construct input sample data, including data acquisition and preprocessing;
[0047] Step 1.1: Obtain the radio frequency signal RF of a single angle plane wave, and perform delay alignment and normalization preprocessing on the radio frequency signal RF, so as to obtain the preprocessed three-dimensional radio frequency data matrix S. In this embodiment, a 2016 The public data set of the annual ultrasonic plane wave challenge, the data includes simulation points, simulation spots, experimental points, experimental spots, and human experiment data, each type of data contains 75 angles of plane wave radio frequency data, read the radio frequency data of a single ultrasonic plane wave, Delay and normalize the radio frequency data of a single plane wave angle to obtain a single ultras...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an ultrasonic plane wave composite imaging method based on deep learning, which comprises the following steps of: 1, constructing input sample data, including data acquisition and preprocessing, 2, constructing mapping from plane wave radio frequency data to an ultrasonic composite image, training a network, and adjusting network parameters, obtaining a planar wave signal, reconstructing an optimal model of a high-quality ultrasonic image; a neural network architecture based on U-Net is specifically adopted, and the whole network comprises a down-sampling module, a middle layer module, an up-sampling module and an output module; and 3, post-processing the image data. According to the method, the ultrasonic image which can only be generated by multiple times of plane wave synthesis is reconstructed from the plane wave signal obtained by single emission through the deep neural network, so that the complexity of obtaining the high-quality ultrasonic image can be reduced, and the frame frequency is improved to promote the development of deep learning ultrasonic imaging.
Description
technical field [0001] The invention relates to the field of image reconstruction of ultrasonic radio frequency data, and specifically relates to a method for reconstructing ultrasonic image of plane wave radio frequency data based on deep learning. Background technique [0002] Ultrasound imaging is a technical method widely used in clinical diagnosis, which has the advantages of high safety, low cost, and non-invasiveness to patients. Plane wave imaging excites all the apertures and emits plane waves from different angles. A set of plane wave echo data is obtained from each angle emission. According to this set of echo data, time-delay superposition is performed, and then envelope and logarithmic compression are performed to obtain an image. Ultrasound image. [0003] Traditional coherent plane-wave compounding (CPWC) synthesizes plane wave data from multiple angles to obtain higher-quality images; minimum variance (MV) beamforming calculates the covariance matrix and inv...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06T11/00G06T5/50G06N3/04G06N3/08A61B8/00
CPCG06T11/001G06T5/50G06N3/084A61B8/52A61B8/5207A61B8/483G06T2207/10136G06T2207/20081G06T2207/20084G06N3/045
Inventor 彭虎谢忠文郑驰超韩志会
Owner HEFEI UNIV OF TECH