Ultrasound Reconstruction Of Image Compression Based On Perception

Ultrasonic testing signal, especially medical ultrasound imaging with narrowed-pulse and very wide frequency band, the higher sampling frequency needed to ensure less loss of information or distortion in echo-RF frequency. Ultrasonic tissue characterization could measure and estimate the characteristics of biological tissue using ultrasonic means and methods. The ultrasound nature of biological tissue could be described by sound reflection, backscatter coefficient, sound attenuation, acoustic nonlinear parameter and other parameters. The main echo signal of the ultrasonic testing was video output and RF output based on analyzing the scattering signal. The most popular RF methods were backscatter integrated and spectrum analysis. The traditional medical ultrasonic imaging need collect and store a amount of RF data, low-efficiency and time-consuming for the RF signal reflecting much information in body tissue.Tao and Donoho proposed compressed sensing theory which breaking through the traditional Shannon sampling theorem, transforming the signal from the high-dimensional to low-dimensional space, then sampling and reconstruction. The compressed sensing theory was used in CT imaging reconstruction by Jakob Heide Jorgensen in2011, the study showed that the dose of X-rays could be reduced by minimizing the pixel. The technology applied to the RF signal detection of ultrasound medical imaging and storage technology could lead to high efficiency for sampling and storing, low-cost, and conducive to online and offline processing for ultrasound RF signal.The paper mainly includes the following aspects:(1) Based on compressed sensing theory, we studied that the applicability of the characteristics of ultrasound image reconstruction by compressed sampling, and studied the sparsity of ultrasound image in time domain and transform domain. We studied the sparse characteristics of ultrasound image by analyzing the coin ultrasound image and kidney analog ultrasound image in time domain and transform domain.(2) We did experiments on coin ultrasound image and kidney simulated ultrasound image in three different ways, firstly, compressed sampling, then reconstruction by L1-Norm. We also studied the advantages and disadvantages of different algorithms used in ultrasound image reconstruction by comparison. (3) We also studied that the OMP algorithm was used in the process of ultrasound image reconstruction through compressed sampling data.We did objective and subject evaluation to ultrasound reconstructed image through reconstructing coin ultrasound image and kidney simulated ultrasound image.We got the following conclusions after our study:(1)After the ultrasound image and kidney coin ultrasound image simulation,the studies showed that, the ultrasonic image was not sparse in time domain, however, the ultrasound images was sparse in the DCT transform domain.(2) OMP algorithm based on compressed sensing techniques in the experiments can be applied in the ultrasound image reconstruction, the choice of the measured value of the reconstructed image affect reconstructed image quality and time,,the quality of reconstructed image get better with the increase of the measured value, but time gets longer.(3)After selecting different DCT coefficient on the characteristics of compressed sensing technology showed that, some of the DCT coefficients selected was good improve the ultrasound image reconstruction, and can greatly reduce the amount of data during sampling. With increased DCT coefficient selected ultrasound image reconstruction quality characteristics getting better, but the time required for the reconstruction process was also increased.Experimental results showed that ultrasound images are not sparse in the time domain, it can be converted into a sparse signal. Compressed sensing technology can be applied to medical ultrasound image reconstruction, greatly reduce the traditional ultrasound image processing heavy amount of data. Currently the technology can continue to strengthen research in this area, so that technology in clinical medicine in the promotion and application of ultrasound. We should continue to do research in the field to promote the technologies in clinical medical ultrasound widely.