Room: AAPM ePoster Library
Purpose: To investigate dosimetric effects of Biozorb markers for proton accelerated partial breast irradiation.
Methods: Two Low-Profile (LP) and four Spiral (SP) markers were selected. Each marker contained 6 titanium clips (clip diameter×width in mm: LP=3×1 and SP=6×1). LP marker sizes (width×length in cm) were 2×3 and 3×3, and SP sizes (diameter×length in cm) were 2×2, 2×3, 3×4, and 4×4. The markers were inserted at the bottom of 5-cm thick breast phantoms (solidifying powder). The phantoms were placed on 8-cm thick of solid water for CT scans. To simulate dose perturbation, a 8×8-cm2 proton beam of 103.23 MeV (range: 8.07 g/cm²) was delivered to the phantoms. For SP2×2 and SP2×3, doses were measured at five depths with a PPC05 chamber with and without clips to determine mass density of clips on CT images in RayStation. EDR2 films were placed 0 and 2 cm below the phantoms and compared to RayStation calculations with and without overriding the density of clips.
Results: Clinically used CT slice thickness of 2.5 mm exaggerated clip sizes and HU values. The chamber measurements revealed the clip density of 1.6 g/cc to produce an acceptable range perturbation in RayStation. Film measurements revealed hot and cold regions surrounding the clips while RayStation (2×2×2-mm³ grid size) only predicted higher dose due to range pullback. Dose differences between film and RayStation at 2-cm depth were slightly greater than those at 0-cm depth. Contouring the clips close to physical size and overriding clip density with 1.6 g/cc (clip artifact with 1.05 g/cc) lowered dose differences (film vs. TPS) by ~2.0%. Dose perturbation by SP was greater than LP at 2-cm depth.
Conclusion: Markers perturbed dose by ~10% more than RayStation could predict. Overriding proper density improved the RayStation dose prediction. The perturbation depended on marker compactness, clip size, and clip orientation.