Proton computed tomography reconstruction using a backprojection-then-filtering approach.

Area of Science:

• Medical Physics
• Imaging Science
• Computational Science

Background:

• Proton computed tomography (CT) is crucial for precise radiation therapy planning.
• Current reconstruction methods face challenges with proton path non-linearity and computational demands.
• Accurate stopping power estimation is vital for effective proton therapy.

Purpose of the Study:

• To introduce a novel backprojection-then-filtering (BPF) algorithm for proton CT reconstruction.
• To develop a list-mode algorithm that accounts for non-linear proton trajectories.
• To enhance the quantitative accuracy and spatial resolution of proton CT imaging.

Main Methods:

• Formulation of a list-mode algorithm for proton CT reconstruction.
• Derivation of an analytical deblurring kernel and a finite matrix correction.
• Utilizing cubic spline fits for estimating proton paths based on entry and exit vectors.
• Validation through Monte Carlo simulations.

Main Results:

• Achieved high quantitative accuracy in relative stopping power (≤0.1%).
• Demonstrated superior spatial resolution compared to standard filtering-then-backprojection methods.
• The BPF algorithm showed computational advantages over iterative reconstruction approaches.

Conclusions:

• The proposed BPF algorithm is a promising candidate for future proton CT systems.
• Its accuracy and efficiency support applications in proton therapy.
• The method effectively addresses challenges of non-linear proton paths and computational load.