Abstract
This paper presents a low-complexity, cost-effective digital beamformer architecture for a high-frequency ultrasound imaging system. The proposed beamformer uses a lookup table and linear interpolation methods for computing the dynamic receive focusing delays and a postfiltering technique to minimize hardware complexity. In the postfiltering technique, channel radio-frequency data having the same fractional delay (i.e., 16f0 resolution) are aggregated prior to interpolation. Thus, only four polyphase structure filters are required in the developed beamformer. In addition, a quadrature bandpass filter that generates an analytic signal is utilized as an interpolation filter; this allows decimation during beam formation and a reduction in computational complexity. The proposed method was evaluated through a 20-μm wire phantom experiment, and the -6-dB lateral and axial resolutions obtained therein were measured and compared with those obtained using a conventional method. The same lateral (165 μm) and axial (80 μm) resolutions at a depth of 5.6 mm were obtained using both the methods, and the proposed method could reduce the beamforming points (i.e., computational complexity) by a factor of the decimation factor (≥4). Images from an excised bovine eye were captured; they showed that the proposed beamformer identified fine anatomical structures such as cornea or iris without compromising the spatial resolution and reduced the computational complexity.
Funding Information
  • 2016 Inje University Research Grant
  • Ministry of Science, ICT and Future Planning, South Korea, through the ICT Consilience Creative Program supervised by the Institute for Information and communications Technology Promotion (IITP-R0346-16-1007)

This publication has 33 references indexed in Scilit: