Efficient and stable beamforming architecture for high frequency ultrasound imaging systems

Abstract

Medical high frequency ultrasound (HFUS) has played an important role in small animal imaging and its application has been extended to clinical use such as diagnosis of musculoskeletal diseases. This is mainly due to its high spatial resolution. To take full advantages of HFUS, full dynamic receive beamforming (DRBF) should be implemented. However, the high sampling rate and computational throughput required in HFUS imaging imposes a huge burden on implementation of the conventional full DRBF algorithm in which beamformation is conducted at all imaging points. This is because the full DRBF generally involves interpolation to increase sampling rate to 16 times the center frequency of transmit ultrasound (f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> ). For example, a 30-MHz ultrasound imaging system should conduct DRBF at the rate of 480 MHz. From this reason, HFUS systems have employed the zonebased DRBF (DRBF-ZB) in which imaging area is divided into several zones and each zone has representative focusing delay. However, this method inevitably suffers from degradation of lateral resolution. In this paper, we propose post-DRBF with polyphase interpolation filter (DRBF-PPF) that is capable of operating at the sampling frequency (i.e., generally 4 times higher than f <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</sub> ). In the proposed method, coarse delays of which resolution is equal to a sampling frequency are applied to RF channel samples at first. For fine interpolation, the coarse delayed samples with identical fractional delay, e.g., 0, 0.25, 0.5, 0.75, are gathered and fed to the corresponding polyphase filters. Final beamformed samples are obtained by summing the output samples of the filters. Note that the processing can be performed at the rate of the sampling frequency.