Efficient Sample Delay Calculation for 2-D and 3-D Ultrasound Imaging.

Area of Science:

• Medical Imaging
• Hardware Acceleration
• Signal Processing

Background:

• Ultrasound imaging is a versatile, cost-effective diagnostic tool.
• Calculating acoustic propagation delays is computationally intensive, especially for 3D systems.
• Efficient computation is vital for low-power, battery-operated ultrasound devices.

Purpose of the Study:

• To explore and evaluate two novel delay generation function designs for ultrasound imaging.
• To quantify the hardware cost, footprint, and performance of these designs on FPGAs.
• To assess the scalability of these architectures for various ultrasound applications, from 2D to 3D.

Main Methods:

• Implementation of two smart delay generation function designs on Field-Programmable Gate Arrays (FPGAs).
• Quantification of hardware cost, footprint, and performance metrics.
• Evaluation of scalability across different ultrasound system complexities (low-power 2D to high-channel 3D).

Main Results:

• Numerical approximation method supports 10,000-channel 3D imaging at 30 fps using 63% of a Virtex 7 FPGA and external memory.
• Exact calculation method achieves 24 fps on 1225-channel 3D imaging with similar FPGA usage and no external memory.
• Both designs are scalable for low-power 2D and ultrafast 2D imaging applications.

Conclusions:

• The proposed FPGA architectures offer efficient solutions for delay generation in ultrasound systems.
• These designs enable high-performance 3D ultrasound and versatile low-power 2D applications.
• The study demonstrates a significant advancement in hardware acceleration for medical ultrasound imaging.