PROCEEDINGS ARTICLE | February 25, 2012
Proc. SPIE. 8320, Medical Imaging 2012: Ultrasonic Imaging, Tomography, and Therapy
KEYWORDS: Signal to noise ratio, Optical filters, Doppler effect, Imaging systems, Ultrasonography, Distortion, Image filtering, Nonlinear filtering, Color imaging, Doppler tomography
In medical ultrasound imaging, dynamic receive beamforming has been used for improving signal-to-noise ratio
(SNR) and spatial resolution. For low-cost portable ultrasound imaging systems, a fractional filter-based receive
beamforming (FFRB) method was previously proposed to reduce the hardware complexity compared to conventional
interpolation filter-based receive beamforming methods (IFRB). While this new beamforming method substantially
reduces the hardware complexity, it yields the nonlinear phase response for high frequencies due to the limited length of
fractional filter coefficients, leading to the bias on flow estimation in ultrasound color Doppler imaging. In this paper, to
evaluate the FFRB method for ultrasound color Doppler imaging, the Field II simulation and string phantom experiments
were conducted. In Field II simulation, the radio-frequency (RF) data were generated by assuming a 7.5-MHz linear
array probe with the transmit frequency of 6 MHz, the ensemble size of 8, and the sampling frequencies of 20 MHz. In
string phantom experiments, the RF channel data were obtained with a commercial SonixTouch ultrasound scanner
equipped with a research package (Ultrasonix Corp., Vancouver, BC, Canada) and a 5-MHz linear array connected to a
SonixDAQ parallel system. The ensemble size and the sampling frequency were set to 10 and 20 MHz, respectively. For
the Field II simulation and string phantom experiments, only 1.2% and 2.3 % in color Doppler estimation error ratio was
observed with mean and standard deviation along the lateral direction. This result indicates that the proposed FFRB
method could be utilized for a low-cost ultrasound color Doppler imaging system with lowered hardware complexity and
minimized phase errors.
