Exact BPF and FBP algorithms for nonstandard saddle curves

Area of Science:

• Medical Imaging
• Computed Tomography

Background:

• Exact cone-beam CT reconstruction from general scanning trajectories is a key research area.
• Nonstandard saddle curve trajectories enable continuous periodic scanning of a volume-of-interest (VOI).

Purpose of the Study:

• To evaluate and compare the performance of filtered backprojection (FBP) and backprojection filtration (BPF) algorithms for cone-beam CT reconstruction.
• To assess reconstruction accuracy and image quality using data acquired along a nonstandard saddle curve trajectory.

Main Methods:

• Implementation of FBP and BPF algorithms in a chord-based coordinate system.
• Utilizing a rebinning procedure to transform reconstructed data into a natural coordinate system.
• Conducting simulations to analyze image quality and noise characteristics.

Main Results:

• The filtered backprojection (FBP) algorithm demonstrated superior image quality compared to the backprojection filtration (BPF) algorithm.
• Both FBP and BPF algorithms exhibited comparable noise characteristics in the reconstructed images.
• Simulation results validated the effectiveness of both algorithms for reconstruction from saddle curve trajectories.

Conclusions:

• Filtered backprojection (FBP) is a more effective algorithm for achieving high-quality cone-beam CT reconstruction from nonstandard saddle curve trajectories.
• Both evaluated algorithms are viable for reconstruction tasks involving continuous periodic scanning, with FBP offering an advantage in image fidelity.