Joint Adaptive Beamforming to Enhance Noise Suppression for Medical Ultrasound Imaging

Abstract

In this work we suggest a combination of two beamformers (BFs) to improve the array noise-suppression abilities using the moments of the eigenvalues (EV) of the data covariance matrix (CM). The eigenspace minimum variance (EMV) BF suffers from the input signal with low SNR, while with high SNR, the dominant mode rejection (DMR) BF degrades. Thus, the random matrix theory (RMT) is used based on the principle that the EV of CM allow predicting the actual moments of the EV so that the SNR level of the proper input data is estimated based on a specified threshold. Compared to the threshold, the higher values of the EV function are associated with the input signal with higher SNR level, so that the EMV BF is adopted, otherwise, the DMR BF. The raw data of the multipurpose phantom (84-317) were acquired using the Verasonics ultrasound system with linear array transducer L11-4v. The performance of the proposed BF (EMV + DMR) was evaluated in terms of lateral resolution using the full width at half maximum (FWHM), peak sidelobe level (PSL) and contrast (CR). Furthermore, the resolution and contrast were improved, indicating that the proposed approach can improve the image quality.