Feasibility study of a compact proton-radiography system based on a scintillator screen

Proton radiography may reduce the range uncertainties affecting proton therapy if it can provide 1% accurate water-equivalent path lengths (WEPL) or relative stopping powers (RSP) of tissues. In a Monte-Carlo study, we showed that such an accuracy can in theory be achieved with a single-2D-detector system. In this follow-up study, we stringently test an experimental implementation of the system using a relatively high (140 MeV) beam energy for samples with 20-mm and 40-mm water equivalence.

The system investigated consists of a scintillation screen coupled with a CCD camera. Samples placed directly upstream of the screen are irradiated with a homogeneous proton field. 2D spatial distributions are obtained of the light emitted (measured) and of the proton fluence (simulated). Their ratio is related to the average proton stopping power distribution in the screen which can be converted to the WEPL distribution of the target. To test the WEPL accuracy, the WEPL and RSP of various homogeneous materials were derived.

The average agreement obtained between the WEPL and RSP derived and the references is 1.6% and 0.7% for the 20-mm and 40-mm water equivalent samples, respectively. Fluctuations in the light output of the screen were identified as the main cause hindering a better agreement.

WEPL determination with the system investigated under stringent test conditions did not result in the desired 1% accuracy. However, optimizing the beam energy with respect to the expected WEPL will likely bring this goal within reach.