A geometric approach to the filtered backpropagation algorithm

Area of Science:

• Medical Imaging
• Computational Physics
• Wave Phenomena

Background:

• Filtered backpropagation is crucial for parallel-beam diffraction tomography.
• Existing methods often rely on Fourier transforms for data processing.

Purpose of the Study:

• To introduce a novel geometric implementation of the filtered backpropagation algorithm.
• To demonstrate its equivalence to established Fourier transform methods.
• To explore potential applications in ultrasound imaging.

Main Methods:

• Geometric implementation of filtered backpropagation using weighted straight-line projections.
• Reorganization of calculations to process plane-wave components separately.
• Theoretical design considerations for an ultrasound scanner.

Main Results:

• The geometric projection method is formally equivalent to the original Fourier transform-based algorithm.
• The approach facilitates separate backpropagation of filtered data's plane-wave components.
• The method suggests a new design paradigm for ultrasound scanners.

Conclusions:

• A geometrically intuitive and computationally equivalent method for filtered backpropagation in diffraction tomography has been developed.
• This approach offers an alternative to Fourier transform methods.
• The findings may pave the way for innovative ultrasound wavefield imaging systems.