Compressed sensing ultrasonic imaging method based on deterministic random measurement matrix

Abstract

The invention relates to a compressed sensing ultrasonic imaging method based on a deterministic random measurement matrix, and belongs to the technical field of ultrasonic imaging. The method comprises the following steps: processing an echo signal received by an ultrasonic array to obtain a required ultrasonic echo signal x; constructing a measurement matrix, namely performing compressed sampling on the ultrasonic echo signal by using the measurement matrix DRS constructed by a deterministic random sequence to obtain a measurement signal y, and performing data transmission; selecting discrete Fourier transform (DFT) as a sparse dictionary psi, and performing sparse representation on the ultrasonic echo signal x; calculating to obtain a coherence coefficient mu between the measurement matrix DRS constructed by the deterministic random sequence and the sparse dictionary psi; and at the receiving end, solving an optimization problem by using a reconstruction algorithm, recovering the ultrasonic signal, and performing beam forming and imaging by using the recovered ultrasonic signal. According to the method, the ultrasonic signal measurement efficiency can be remarkably improved, and the ultrasonic signal can be quickly and approximately optimally reconstructed.

Description

technical field [0001] The invention belongs to the technical field of ultrasonic imaging, and relates to a compressive sensing ultrasonic imaging method for constructing a measurement matrix from a deterministic random sequence. Background technique [0002] With the improvement of the quality requirements of ultrasonic imaging, a higher sampling frequency is required, so the amount of echo data to be stored is very large, which increases the complexity of hardware implementation. Compressed sensing theory (CS) is a method proposed for high-speed data acquisition and large-capacity data storage in recent years. The theory believes that when the signal itself or in a certain transform domain is sparse, it can be much lower than the Naiqui Ster's rate compresses the signal while sampling, and then through the reconstruction algorithm, the original signal can be reconstructed from a small amount of sampled data with extremely high precision, reducing the amount of data that ne...

AI Technical Summary

Problems solved by technology

This type of matrix has a simple structure and good reconstruction effect, but when the number of measurements is small, it is still necessary to construct a high-dimensional matrix first, and then select rows, which wastes storage resources and has not been widely used.

In particular, image reconstruction is poor when these measurement matrices are applied to overlapping ultrasound signals

Images