MO-FG-CAMPUS-JeP1-04: Evaluating DECT Vs SECT Range Differences in Proton Therapy Using Clinical Data.

PURPOSE: Range uncertainty from X-ray CT number conversion to stopping power ratio (SPR) is one of the key factors limiting the potential of proton therapy. The large margins required for deep seated tumors degrade the organ sparing achievable with the technology. Of interest is the application of dual energy CT (DECT) to SPR estimation. In this planning study proton range differences between SECT and DECT have been quantified for brain cases.

METHODS: A last generation dual source DECT scanner was used to acquire SECT (150 kVp with Sn filtration) and DECT (additionally 90 kVp) scans of phantoms and 5 head trauma patients, acting as surrogate cancer patients. Phantom materials were characterized in terms of SPR in a particle beam to obtain reference values. IMPT treatment plans were generated on the basis of SECT and DECT SPR images for hypothetical brain tumors using a short and a long beam path. Range differences between SECT and DECT from plan recalculations were evaluated in beam-eye-view (BEV) by comparing the 80% isodose.

RESULTS: For the 18 phantom materials the SECT RMS SPR errors were 2.6% compared to 1.1% for DECT. Group median relative range differences between SECT and DECT plans were -1.0% for the short beam path over the 5 patients investigated in this study. For the long beam path the median difference was -1.4%. These relative range differences corresponded to -0.5 mm and -1.4 mm shifts respectively.

CONCLUSION: This is the first study performing proton therapy treatment planning on DECT patient images. Important range differences of more than 1 mm were observed between SECT and DECT treatment plans, and DECT SPR accuracy was found superior on the basis of phantom measurements. While the patients investigated in this study did not have brain tumors, the findings we observed should apply to cancer patients. Deutsche Forschungsgemeinschaft (MAP); Bundesministerium fur Bildung und Forschung (01IB13001).