Studying the theoretical limits of a semi-analytical WEPL estimation algorithm for Sandwich-TOF iCT

Abstract

Ion beam therapy is a precise treatment modality for deep-seated tumors, relying on accurate determination of the relative stopping power (RSP) distribution within patient tissue. Due to conventional X-ray computed tomography (xCT) measuring the interaction of photons with tissue, RSP conversion errors emerge from linear photonic attenuation to ionic stopping power of 1.6-5%, potentially affecting treatment accuracy. Alternatively, ion computed tomography (iCT) offers a direct RSP measurement, mitigating these uncertainties. Current iCT systems consist of tracking detectors and separate calorimeters and can achieve RSP errors on the order of 0.5%. While these types of iCT systems are promising, a large challenge these systems face is reaching data acquisition rate sufficiently high enough to complete a clinical scan within minutes. This thesis investigates the Sandwich time-of-flight ion computed tomography (TOF-iCT) approach, which determines the energy loss of the particles via a time-of-flight (TOF)-based water-equivalent path length (WEPL) estimation. In contrast to other ion imaging concepts, where the energy loss is estimated via a separate calorimeter downstream of the patient, this method does not require a residual energy detector as it determines the WEPL via a TOF measurement through the patient. Consequently, a much more compact scanner can be built, which could be more cost-effective and easier to integrate into the treatment room. Recent studies have investigated the feasibility of this new imaging method using a simplified WEPL estimation algorithm. However, experiments as well as simulations of this modality have shown significant systematic errors, highlighting the need for a more refined WEPL estimation approach. To enhance the WEPL estimation accuracy, this work investigates a semi-analytical approach with Geant4 simulations. These simulations are intentionally idealized and serve to explore the theoretical limits of the proposed method, while also demonstrating its compliance with the technical requirements of iCT systems. Monte Carlo methods were employed to analyze the impact of sampling width of the particle’s step and trajectory, beam energy, and material properties on WEPL accuracy. The method was validated using experimental data from Med Austron. It was shown that this new approach is a clear improvement to the previous calibration-to-air and a root-mean-square relative difference to ground truth for the WEPL of less than 0.5% could be achieved for all materials tested. Additionally, potential points of improvement have been identified. The findings contribute to the optimization of TOF-iCT for improved RSP estimation, potentially advancing the accuracy of ion beam therapy treatment planning.