Sensitivity study for CT image use in Monte Carlo treatment planning

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Dosimetry

Background:

• Monte Carlo treatment planning (MCTP) relies on accurate patient data segmentation into voxel phantoms for dose calculation.
• Conventional treatment planning (TP) assigns mass densities, while MCTP also requires accurate media assignment.
• Incorrect media assignment in CT segmentation can introduce substantial dose calculation inaccuracies in MCTP.

Purpose of the Study:

• To investigate the impact of CT segmentation errors on Monte Carlo treatment planning (MCTP) dose calculations.
• To quantify the inaccuracies in MCTP arising from mis-assigned media in voxel phantoms.
• To emphasize the importance of precise CT calibration using appropriate phantoms.

Main Methods:

• Utilized a test phantom with a precisely known composition to simulate CT segmentation.
• Performed dose calculations using Monte Carlo methods based on segmented voxel phantom data.
• Analyzed dose errors across different radiation types (photons and electrons) and energies.

Main Results:

• Observed significant dose errors in specific regions due to CT segmentation inaccuracies.
• Documented dose errors up to 10% for 6 and 15 MV photons.
• Quantified substantial errors exceeding 30% for 18 MeV electrons and 40% for 250 kVp photons.

Conclusions:

• A meticulous CT calibration process utilizing a suitable phantom is essential for accurate MCTP.
• Generic CT calibrations and the use of commercially available CT phantoms require critical evaluation.
• Addressing CT segmentation accuracy is vital to minimize dose errors in radiation therapy.