Feasibility study of a proton CT system based on 4D-tracking and residual energy determination via time-of-flight

TL;DR

This paper proposes a novel proton CT system using 4D-tracking LGAD detectors and time-of-flight energy measurement, demonstrating high RSP accuracy and potential for improved ion beam therapy planning.

Contribution

It introduces a new pCT system based on LGAD detectors and TOF measurement, with optimized design parameters and calibration for enhanced RSP accuracy.

Findings

RSP accuracy < 0.6% achieved in simulations

Design parameters significantly impact RSP resolution

Calibration procedure improves measurement precision

Abstract

For dose calculations in ion beam therapy, it is vital to accurately determine the relative stopping power (RSP) distribution within the treated volume. Currently, RSP values are extrapolated from Hounsfield units (HU), measured with x-ray computed tomography (CT), which entails RSP inaccuracies due to conversion errors. A suitable method to improve the treatment plan accuracy is proton computed tomography (pCT). A typical pCT system consists of a tracking system and a separate residual energy (or range) detector to measure the RSP distribution directly. This paper introduces a novel pCT system based on a single detector technology, namely low gain avalanche detectors (LGADs). LGADs are fast 4D-tracking detectors, which can be used to simultaneously measure the particle position and time with precise timing and spatial resolution. In contrast to standard pCT systems, the residual energy is determined via a time-of-flight (TOF) measurement between different 4D-tracking stations. The design parameters for a realistic proton computed tomography system based on 4D-tracking detectors were studied and optimized using Monte Carlo simulations. The RSP accuracy and RSP resolution were measured inside the inserts of the CTP404 phantom to estimate the performance of the pCT system. After introducing a dedicated calibration procedure for the TOF calorimeter, RSP accuracies < 0.6 % could be achieved. Furthermore, the design parameters with the strongest impact on the RSP resolution were identified and a strategy to improve RSP resolution is proposed.