Theoretical tumor edge detection technique using multiple Bragg peak decomposition in carbon ion therapy

Abstract

Purpose: The range precision in carbon ion therapy is extremely sensitive to tissue density variations. A high energy carbon beam, after crossing a high-gradient edge parallel to the beam direction, suffers from range mixing leading to the detection of multiple Bragg peaks (BPs) varying intensity and water equivalent thickness (WET). The purpose of this work was to introduce a model that determines the position of a high-gradient edge based on information acquired from carbon transmission imaging. Methods: A model was derived to determine the lateral distance between the irradiation beam propagation axis and the edge position. To validate it, carbon beams were simulated and propagated through two parametric phantoms: (1) a bone cube in a water tank and (2) a semi-cylindrical bone insert in a water tank. The method was tested in a lung tumor case where range mixing led to more than two BPs being detected, requiring an iterative BP decomposition to determine the fraction of carbon ions crossing the materials surrounding the edge of interest. Results: The theoretical model predicted the edge position relative to the beam position with an error ≤1 mm for all studied cases with a maximum dose delivered of 24 μGy. Conclusions: The method presented here is a proof of principle. It does not take into account clinical uncertainties. However, this approach provided promising results suggesting that future extension to assess the impact of clinical uncertainties should be performed.