Adaptation of Range-Doppler Algorithm for Efficient Beamforming of Monostatic and Multistatic Ultrasound Signals

Area of Science:

• Ultrasound imaging
• Signal processing
• Medical imaging technology

Background:

• Conventional delay-and-sum (DAS) beamforming is slow for large datasets.
• F-k migration speeds up synthetic aperture (SA) ultrasound but is memory-intensive and limits parallel processing.

Purpose of the Study:

• To introduce and evaluate the range-Doppler algorithm (RDA) for SA ultrasound beamforming.
• To compare RDA's performance against DAS and F-k migration in terms of speed, resolution, and memory usage.

Main Methods:

• Simulations and phantom experiments were used to assess RDA performance.
• RDA was implemented using a spatial Fourier transform, adapted from synthetic aperture radar (SAR) imaging.
• Axial sidelobes in RDA images were mitigated using temporal frequency binning.

Main Results:

• RDA achieved comparable lateral resolution and contrast to DAS and F-k migration.
• RDA processing time was ten times faster than DAS on a single CPU and required half the memory of F-k migration.
• RDA supported significantly more parallel threads (1000+) than F-k migration (8).

Conclusions:

• RDA presents a computationally efficient and memory-sparing alternative for SA ultrasound beamforming.
• RDA's suitability for parallel processing makes it advantageous for graphics processing units (GPUs) and cost-effective ultrasound devices.
• RDA offers a substantial speed advantage for devices with moderate parallel processing capabilities.