The Research On Beamforming Algorithm Based On Plane-wave Transmission

Medical Ultrasound Imaging,Computed Tomography,and Magnetic Resonance Imaging are the three most commonly used clinical imaging techniques.However,medical ultrasound imaging has been widely used in clinical practice because of its characteristics of no radiation,portability,real-time imaging,and relatively low price.In ultrasound imaging systems,beamforming technology plays a decisive role in image quality.The beamforming methods currently widely used in the business community is delay and sum(DAS)combined with dynamic aperture apodization technology based on the line-by-line scanning modality.However,the line-by-line scanning modality severely limits the frame rate and gradually becomes inapplicable in certain scenarios with the required frame rate relatively high.The scanning modality based on plane wave transmission only emits one plane wave to demonstrate the entire imaging area,which can achieve a very high frame rate.However,the accompanying problem is that the quality of the image is greatly degraded.To solve this problem,coherence compounding and adaptive beamforming were introduced.The coherent plane-wave compounding technology achieves a compromise between frame rate and image quality.Obviously,when adopting a small number of plane-waves,the frame rate will be greatly improved,but the image quality will also be degraded.The decrease in amplitude,especially the suppression of artifacts caused by the side lobes.Additionally,it has been reported there is also other artifacts in CPWC known as axis artifacts.this paper focuses on how to apply the image optimization method in aperture domain to further suppress the side-lobe level to improve image quality and to try to eliminate or suppress such artifacts according to the generation mechanism of the axial lobe.Adaptive beamforming can be divided into two major categories: minimum variance lossless response(MV)and coherence factor(CF).If it is directly applied to the plane-wave imaging,the resulting image will show a decrease in the brightness in all imaged region,the speckle pattern will become uneven,and even the black band artifacts which appearing in the imaged region surrounding the highlight imaged objects.MV based on the eigen-space(ESBMV),what the corresponding eigenvectors with larger eigenvalues are selected to form desired signal subspace and the remaining components of eigenvectors to construct the noiseinterference signal subspace by fault,is completely feasible in scenarios with high signal-tonoise ratio.However,when the signal-to-noise ratio or signal-to-noise noise is low,such as in the region near the bright imaged objects,the side-lobe level is much larger than the real signal level.Therefore,this paper improves classification method of signal subspace and adds the criterion that the angle between the steering vector and the eigenvector to suppress the artifact caused by the misclassification in original ESBMV.In the coherence factor,although many scholars have modified the original CF,such as the phase/symbol coherence factor,scalar wiener factor,general coherence,spatio-temporal smoothing coherence factor and so on.It can be clearly seen that above four methods all introduce an adjustment factor to achieve a compromise.The condition of each imaging point is not the same,and it is obviously that it is not sensible to take a fixed adjustment value in every imaged object.Therefore,in this paper,we evaluate the characteristics of each imaged point and map the adjustment value to each imaging point to further improve the imaging performance.In summary,this paper focuses on the research of beamforming technology based on planewave scanning modality,studies on apodization,coherence assessment,optimal MV weight vector in steering angle domain to further suppress the side-lobe level and improve image quality,and suppressing axis artifacts according to the formation mechanism.In the singleangle plane wave,the adjustment factors dependent with the local SNR are used to optimize the performance of the algorithm.In the ESBMV,the classification criteria for the desired signal space and the noise-disturbance signal space are modified to make it more accurate.Finally,the nature of the ESBMV algorithm is further studied,and the high performance nature of the algorithm is proposed as the cyst area of the test subject which intensity is round zero and artifact overlap coincidentally created,and a series of hypoechoic cysts were designed to test the conjecture proposed in this paper.