Ultrasound Transcranial Imaging Based on Fast Coherent-time-delay and Correlative Pixel-based Beamforming

Chen Bai,Xinyu Zhang,Xiaoyang Qiao,Yuchao Sang,Hui Zhong,Mingxi Wan
DOI: https://doi.org/10.1109/ultsym.2018.8579926
2018-01-01
Abstract:Adaptive beamforming, such as the fast minimum variance (FMV) algorithm, has been developed to improve the contrast-to-noise ratio (CNR) in plane wave imaging with extremely reduced computation. However, in ultrasound transcranial imaging, FMV will be confronted with challenges raised by echoes with low intensities, especially when calculating the beamformer contributions at points which are in a distance from the beam center line. In order to improve the CNR and resolution in low echo beamforming during transcranial imaging, this paper proposed an adaptive beamforming which combined with fast minimum variance (FMV) algorithm and unified pixel-based beamforming, i.e. fast coherent-time-delay and correlative pixel-based (FCCP) beamforming. By considering the influence of neighboring pixels during the beamforming, this method can remove undesirable information instead of increasing diagnostic information in transcranial imaging. The influence of neighboring pixels was reflected by coherent-time-delay and coherent coefficients of each covariance matrix. After the simulation and phantom study, the results prove that CNR and the resolution are indeed ameliorated be FCCP, compared with delay-and-sum (DAS) and conventional minimum variance (MV) algorithm. In simulation, the CNR was augmented by 6 dB as well as the lateral resolution was improved by from 0.1536 mm from 0.1740 mm. FCCP is the combination of compound coherent-time-delay and correlative pixel-based, which could improves the resolution and contrast if image, and FMV. FMV is involved for the high complexity caused by matrix inversion, the computation was simplified by principal component analysis (PCA) when calculate the covariance matrix.
What problem does this paper attempt to address?