An optimization algorithm for dose reduction with fluence-modulated proton CT.

Area of Science:

• Medical Physics
• Radiological Imaging
• Radiation Oncology

Background:

• Proton computed tomography (pCT) utilizes pencil beam scanning for imaging.
• Current pCT methods often use uniform proton fluence, which may not be dose-optimal.
• Task-specific noise reduction requires modulation of proton fluence.

Purpose of the Study:

• To present a fluence field optimization method for FMpCT.
• To investigate dose reduction performance for various phantoms and image variance targets.
• To assess the feasibility of achieving prescribed image variance distributions.

Main Methods:

• Monte Carlo simulations of a pCT prototype scanner were used to estimate variance.
• An iterative approach calculated target variance projections for filtered backprojection reconstruction.
• Pencil beam weights were determined by fitting a model to variance ratios.

Main Results:

• Dose reduction of 8.9% (water phantom) and 16.0% (head phantom) was achieved for constant image variance.
• Targeted ROI imaging yielded dose reductions of 25.7%–40.5% outside ROIs.
• Doses inside ROIs increased by 9.2%–19.2% but were acceptable if aligned with therapeutic regions.

Conclusions:

• A novel fluence field optimization method for FMpCT was developed.
• The method successfully achieved prescribed image variance targets.
• FMpCT offers considerable dose reduction, beneficial for image guidance and adaptive therapy.