Rectified wavenumber algorithms for efficient ultrasound imaging

Area of Science:

• Medical Imaging
• Ultrasound Technology
• Computational Imaging

Background:

• Wavenumber-domain beamforming offers computational benefits for ultrasound image formation.
• Existing wavenumber-domain methods can cause distortions at large angles and depths, limiting their clinical application.
• Conventional delay-and-sum (DAS) beamforming provides high-quality images but is computationally intensive.

Purpose of the Study:

• To develop a wavenumber-domain beamforming framework that is consistent with DAS beamforming.
• To retain the computational efficiency of Fourier-domain processing while improving image quality.
• To create a versatile algorithm applicable to various ultrasound imaging schemes.

Main Methods:

• A novel wavenumber-domain framework was developed, building on DAS-consistent imaging principles.
• The method incorporates modified Fourier-domain weighting and axial scaling to correct for spatial-frequency filtering.
• The algorithm was validated using simulations, phantom experiments, and in-vivo liver imaging data.

Main Results:

• The proposed method produced images with DAS-equivalent amplitude, speckle statistics, resolution, and contrast.
• Structural Similarity (SSIM) values between the proposed method and DAS exceeded 0.99.
• The new method demonstrated a significant reduction in computational cost (approximately one order of magnitude) compared to conventional DAS.

Conclusions:

• The developed wavenumber-domain framework achieves DAS-consistent ultrasound image formation with high computational efficiency.
• This method overcomes limitations of previous wavenumber-domain approaches, offering improved image fidelity.
• The algorithm is suitable for real-time and resource-constrained ultrasound systems, advancing medical imaging capabilities.