Quantitative Framework for Adopting Dect with Dspr into a Proton Clinic
Abstract
Purpose
In radiation oncology, advanced imaging tools such as dual-energy computed tomography (DECT) offer the potential for improved tumor delineation and reduced proton range uncertainty. Such technologies are often introduced into clinical workflows without comprehensive validation across scanner platforms or clearly defined quality assurance strategies. This work presents a systematic framework for validating DECT-derived direct stopping-power ratio dSPR, evaluating its reproducibility, and assessing its potential clinical impact in proton therapy.
Methods
Validation was performed using commercial tissue-equivalent phantoms and in-house–developed phantoms representing clinically relevant materials (bone, fat, water, lung). Identical DECT acquisition and reconstruction protocols were applied across three CT scanners from a single vendor. Inter-scanner reproducibility was assessed by comparing dSPR-derived proton range metrics (R90) in the treatment planning system. One purported benefit of dSPR is that it presents a scanner independent stopping power ratio curve for proton therapy planning. To evaluate dosimetric implications, a simplified toy-model was implemented within the treatment planning system. Proton treatment plans were generated using identical beam geometry while systematically varying assumed range uncertainty values. Dose–volume metrics for the target and adjacent organs-at-risk were evaluated.
Results
R90 agreement across scanners was within approximately 1 mm/1% for soft-tissue and adipose-equivalent materials, while larger variability was observed for higher-density and calcium-based materials, indicating residual scanner dependence in dSPR assignment. In the toy model planning study, reduced range uncertainty resulted in improved sparing of organs-at-risk distal to the target while maintaining target coverage.
Conclusion
Phantom-based validation demonstrated material- and scanner-dependent variability in DECT-derived dSPR that should be characterized prior to clinical use. Variability in scanner performance may also have implications for multi-institutional studies. Dosimetric evaluation suggested that validated reductions in range uncertainty may translate into improved organ-at-risk sparing in proton therapy. Routine phantom-based quality assurance is recommended beyond initial commissioning to confirm consistency over time.