A simulation study on proton computed tomography (CT) stopping power accuracy using dual energy CT scans as benchmark

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Imaging Science

Background:

• Accurate stopping power estimation is critical for proton therapy, with current uncertainties being the primary source of dose margin expansion.
• Dual-energy CT and proton CT are imaging modalities proposed for generating patient stopping power maps.
• Current methods have limitations impacting the precision of proton therapy dose calculations.

Purpose of the Study:

• To evaluate the accuracy of proton CT for stopping power estimation.
• To compare proton CT accuracy against dual-energy CT using phantom data.
• To establish reference accuracy levels for proton CT in a controlled setting.

Main Methods:

• Dual-energy and single-energy CT scans were performed on calibration and abdomen phantoms.
• Proton CT scans were simulated using Monte Carlo methods, incorporating realistic detector noise.
• Stopping power maps were computed for all scan types and compared to ground truth values.

Main Results:

• Proton CT demonstrated superior stopping power estimation accuracy, with lower root mean square errors (0.2%-0.5%) compared to dual-energy CT (0.5%-0.9%).
• Single-energy CT showed significantly higher root mean square errors (1.6%-2.7%).
• Maximal errors were substantially reduced with proton CT (0.51%) versus dual-energy CT (1.7%) and single-energy CT (7.4%).

Conclusions:

• Improved stopping power estimation is essential for minimizing range errors in proton therapy.
• Proton CT shows promise for achieving the necessary advancements in accuracy.
• Further development of proton CT could lead to more precise and effective proton therapy treatments.