BLUE RIBBON POSTER THERAPY: Clinical Dosimetric Impact of Dual-Layer Dect–Derived SPR for Proton Therapy Planning
Abstract
Purpose
To assess dual-layer DECT-derived SPR images for proton therapy by identifying tissue-specific deviations versus SECT, validating discrepancies using measured SPR in biological tissues, and evaluating dosimetric impact of DECT-based robust optimization.
Methods
SECT and DECT images from the CIRS-62M phantom and 24 patients were imported into the TPS. SPR agreement was evaluated using a 2D joint voxel distribution stratified into lung, fat, muscle, and bone. Six biological tissue samples (beef/pork) were included. Reference SPR for phantom plugs and tissues was measured using a 4-mm collimated 200-MeV proton beam with a multi-slit ion chamber (MSIC) and compared to SECT/DECT estimates. Differences were quantified in shared ROIs. SECT-clinical plans (3.5% range, 3/5 mm setup) were recalculated and reoptimized on DECT using 2.5% range uncertainty, and OAR dose impact was assessed for targets abutting critical OARs.
Results
SPR measurements of the non-biological CIRS-62M plugs (which were included in the SECT-derived calibration curve) agreed with DECT predictions by 5.28% on average and with SECT predictions by 4.64%. The largest WET deviations occurred in bone plugs where DECT/SECT deviations were −2.09%/6.5%. In six biological bone samples, DECT SPR agreed with MSIC reference values within 0.9–6.7% (mean 3.1%) versus 0.24–9.0% (mean 3.6%) for SECT. Across patients, SECT–DECT OAR SPR differences averaged 0.93%, highest in lung (3–9%) and bone (~2%). Dose re-computation on DECT showed <1% OAR and target differences. For selected cases, DECT reoptimized plans (2.5% range uncertainty) reduced OAR dose by up to 3 Gy (hippocampi mean) and ~1 Gy (esophagus/brainstem mean).
Conclusion
DECT provided improved agreement with measured SPR in biological tissues where SECT stoichiometric calibration shows larger deviations, particularly in lung and bone. DECT-based planning may support reduced range uncertainty and improved healthy tissue sparing without compromising target coverage.