ESTRO 37 Abstract book

ESTRO 37

S465

S. Moragues Femenia 1 , J.F. Calvo-Ortega 1 , J. Casals 1 1 HOSPITAL QUIRON BARCELONA, Radiotherapy, Barcelona, Spain

Purpose or Objective Most dosimetric quality assurance (QA) devices have low spatial resolution. In treatments with small or modulated fields such as intensity modulated radiotherapy (IMRT), stereotactic body radiation therapy (SBRT) and stereotactic radiosurgery (SRS), a high resolution dose measurement technique would be advantageous, and radiochromic film dosimetry is suitable for this purpose. In this work, two different multichannel film dosimetry algorithms using the triple channel approximation and a hybrid version have been compared. Material and Methods EBT3 film was used for QA of IMRT and SBRT cases for a range of anatomical sites. Dose distributions were generated from the film using: 1. Red channel only 2. Méndez multichannel algorithm using the truncated normal distribution (TN) [1] 3. Pérez-Azorín multichannel algorithm [2] 4. A hybrid of the Méndez and Pérez-Azorín algorithms developed in his study. All images were analysed at the same resolution. An adaptive median filter was used to minimise noise dependencies and the calibration curve was fitted to a rational parametric form. To perform absolute dosimetry, net optical density was used rather than pixel value. OmniPro I’mRT software was used to perform absolute gamma analysis using 3%/3mm and 2%2mm criteria on a 20% threshold comparing the various generated distributions from film to the calculated distribution from the treatment planning system[M1] (TPS) (Monaco v5, Elekta). All plans were calculated in the same TPS Results Figure 1 shows a representative dose profile comparison from a bilateral breast case comparing the hybrid algorithm (4) from EBT3 film with the TPS. Table 1 depicts the gamma analysis results for a range of IMRT and SBRT cases. Dose delivered ranged from 180 cGy to 1800 cGy.

Purpose or Objective To evaluate the accuracy of DVH calculate against published analytical values. Material and Methods

ion

The designed datasets and metrics reported by Nelms et al (Med Phys. 2015 Aug;42(8):4435-48) have been use as reference (“ground truth”) to check the accuracy of the DVH calculated by Eclipse v 10.0. These datasets include DICOM RT structure sets (a small sphere, cylinder, and cone) created programmatically with axial plane spacing varying from 0.2 to 3 mm, and synthetic DICOM RT dose files generated with 1D linear dose gradient and with grid resolution varying from 0.4 to 3 mm. We focused our investigation for the scenario of CT scans with 1 mm axial spacing and 1 mm calculation grid size, as used routinely in our cranial stereotactic (SRS) plans. DVHs for 20 combinations RT structure/RT dose file were analyzed. Differences (Eclipse vs. ground truth) in the following metrics were calculated: structure volume (V), Dmax, Dmin, D99%, D95%, D5%, D1% (where Dx% is the dose to x% volume), and D0.03 cm 3 (minimum dose at 0.03 cm 3 ). Differences are referred as local percent differences, i.e, all normalized based on the ground truth values reported by Nelms et al. Results Table 1 shows the differences obtained. Differences on volume metric were within 2.5%. Very large differences (up to 70%) were found on Dmin, which were reduced within 10% for D99%. Better agreement (within 5%) was noted for D95%. Differences on remaining metrics were within 5%.

Figure 1: Dose profile of a bilateral breast case comparing QA dose calculation from the Monaco TPS (green) with the measured dose determinad from and EBT3 film using the hybrid algorithm (4)(red ).

Conclusion In SRS plans are commonly normalized such as D99% in target volume is equal to the prescribed dose. As large differences (up to 10%) have been noted on D99%, it seems reasonable to normalize the SRS plan using the D95% metric instead of D99%. PO-0881 Absolute dosimetry with EBT3 films F.J. San Miguel Avedillo 1 , F. Clemente Gutierrez 1 1 Hospital de la Defensa Gomez Ulla, Radiotherapy, madrid, Spain