Robust range prediction for arbitrary tissue mixtures based on dual-energy CT


Robust range prediction for arbitrary tissue mixtures based on dual-energy CT

Möhler, C.; Wohlfahrt, P.; Richter, C.; Greilich, S.

The treatment planning of proton or ion radiation therapy is affected by uncertainties arising from the heuristic conversion of computed tomography (CT) images to stopping-power ratio (SPR) maps. In this work, we present how these uncertainties can potentially be reduced by the use of dual-energy CT (DECT), via a physics-based SPR prediction. According to the Bethe formula, the SPR is the product of the electron density and the stopping number relative to water. The latter ranges between 0.96 and 1.02 for human tissue at a therapeutic beam energy of 200 MeV/u and depends on the mean excitation energy (I-value).
As a first step, the relative electron density can be directly determined from DECT images in a universal and robust procedure, based on a simple assumption for the cross section parameterization. Secondly, we propose to infer the relative stopping number from the relative photon absorption cross section obtained from DECT scans - instead of using an effective atomic number as a proxy for the I-value, which has previously been suggested in literature. Our choice of variables makes a proper treatment of tissue mixtures possible, which inevitably occur in patient CT images, and allows for a convenient definition of the uncertainties.
A calculation-based analysis of tabulated body tissues and tissue base components - such as water, lipid, carbohydrates and protein - suggests a maximum uncertainty below one percent for arbitrary mixtures of human tissue. We performed first experiments, combining particle range measurements with DECT scans, to validate our method of stopping-number prediction.

Keywords: dual-energy CT; proton therapy; ion-beam therapy

  • Poster
    55th Annual Conference of the Particle Therapy Co-operative Group (PTCOG), 22.-28.05.2016, Prag, Czech Republic

Permalink: https://www.hzdr.de/publications/Publ-23909
Publ.-Id: 23909