Compressed sensing approach for CMUT sparse array in multi-element synthetic transmit aperture

Abstract

In order to improve the imaging quality of capacitive micromachined ultrasonic transducer (CMUT) sparse array, the authors propose an imaging strategy combining multi-element synthetic transmit aperture (MSTA) imaging mode and compressed sensing (CS) framework. The imaging strategy consists of four steps: First, use the genetic algorithm (GA) to sparse the CMUT array and obtain the position of the receiving active array elements. Second, the CMUT sub-arrays are used to transmit ultrasonic signal in turn, and the sparse array obtained in the first step is used to receive ultrasonic echoes. Third, the received under-sampling channel data are used to recover the full channel data with the CS reconstruction algorithm. Finally, an MSTA image is beamformed from the recovered full channel data. This imaging strategy is called MSTA-CS mode. Due to the efficient sparse recovery capability of CS technology, the MSTA-CS imaging strategy not only reduces the number of transmitting hardware channels, but also improves the imaging clarity of CMUT sparse array. Both the Field II simulation and the breast phantom experiments demonstrate that the MSTA-CS mode achieves higher contrast and structural similarity under the same receiving sparse array conditions. • In this paper, Multi-element Synthetic Transmit Aperture (MSTA) method is used in ultrasound imaging of CMUT sparse array, this method reduces the number of transmitting hardware channels. • Under the condition of lateral under-sampling of CMUT sparse array, the closed array elements echoes data are recovered by using compressed sensing framework. • Take the full array imaging results as the standard, and verify the compressed sensing theory through breast phantom experiments.