Software-based Processing for Contrast-Enhanced Ultrasound Imaging using Pulse-Inversion Spectral Deconvolution

A software-based processing approach for contrast-enhanced ultrasound (US) imaging is presented. Termed pulse-inversion spectral convolution (PISD), the main motivation here was to further evaluate the performance of this contrast-enhanced US imaging technique using a series of experimental studies. The PISD-based contrast-enhanced US approach is founded on a class of Gaussian derivative functions (GDFs). Two types of PISD pulses (one GDF and convolution of three GDF pulses) can be used to form two new inverted pulse sequences. These pulses are then used to filter backscattered US data for isolation of the nonlinear MB signal component. An US system (Vantage 256, Verasonics Inc) with a L11-4v linear array transducer was programmed and used for PISD-based contrast-enhanced US imaging. The receive data from all channels was shaped using a wide beamforming technique. In vitro US images were collected from a flow phantom perfused with a MB contrast agent using PISD and traditional pulse-inversion US (nonlinear, NL) for comparison. Further, role of the transmit aperture size (
$\text{TX}=32$
or 64) was also evaluated using a US pulse frequency of 6.25 MHz and a low mechanical index of 0.2. Contrast enhancement was measured using a contrast-to-noise ratio (CNR). Preliminary in vivo data was acquired from the hindlimb (femoral artery) of healthy rats after receiving a bolus injection of MBs. Overall, widebeam contrast-enhanced US imaging using the larger aperture size of 64 elements yielded improved CNR values. Using this widebeam US imaging approach, both in vitro and in vivo results demonstrated that the PISD-based contrast-enhanced US technique produced images with improved contrast compared to the more traditional B-mode and NL US imaging strategies.