A fully programmable computing architecture for medical ultrasound machines

Area of Science:

• Medical Imaging
• Digital Signal Processing
• Computer Engineering

Background:

• Traditional medical ultrasound systems rely on application-specific integrated circuits (ASICs) for high computational demands in receive beamforming.
• Developing flexible and efficient processing architectures is crucial for advancing ultrasound technology.

Purpose of the Study:

• To present a simple, programmable computing architecture for core ultrasound signal processing.
• To evaluate the feasibility of using a Field-Programmable Gate Array (FPGA) and Digital Signal Processor (DSP) for medical ultrasound applications.

Main Methods:

• Developed a programmable architecture integrating an FPGA and a DSP.
• Utilized previously established efficient front-end algorithms for ultrasound signal processing.
• Assessed resource utilization for B-mode imaging with 64 channels and 120 scanlines/frame at 30 frames/s.

Main Results:

• The proposed architecture required 97.3% of FPGA resources and 51.8% of DSP resources for comprehensive front-end and back-end processing.
• Demonstrated the capability to support B-mode imaging at 30 frames per second with specified channel and scanline configurations.

Conclusions:

• The programmable FPGA-DSP architecture is suitable for low- and medium-level ultrasound machines.
• This architecture offers a flexible platform for developing and deploying new algorithms and clinical applications in medical ultrasound.