ESTRO 2021 Abstract Book

S1569

ESTRO 2021

Conclusion Preliminary results demonstrate that dose constraints can be retrieved through a non-linear embedding of kNN treated patients with similar anatomy and prescription and their dose constraints. Harnessing these constraints could be used to provide end-to-end dose planning and deliver clinically acceptable plans. Large scale tests are ongoing involving cohorts with more patients and additional anatomies. PO-1840 A planning study evaluating the Advanced Collapsed cone Engine in HDR skin brachytherapy R. Stansbridge 1 , R. Caines 1 , C. Lee 1 1 The Clatterbridge Cancer Centre, Physics, Liverpool, United Kingdom Purpose or Objective Patient dosimetry in Brachytherapy is usually determined under the TG-43 formalism; this calculation is based on data from an Iridium-192 source dwelling at a single position in the centre of a large volume of water. The Advanced Collapsed cone Engine (ACE) within the Oncentra Brachy TPS (Elekta, Sweden) is a model-based dose calculation algorithm developed to compensate for the dose calculation conditions disregarded in TG-43 calculations. For HDR brachytherapy skin treatments, the presence of a tissue-air interface is not considered in TG-43 calculations which can potentially result in a reduced dose delivered due to the missing scattered photons. In current local practice water equivalent bolus is added on top of the patient during treatment to simulate the TG-43 dose calculation conditions. A planning study was conducted by comparing ACE with TG-43 for 27 previously treated HDR skin treatments against ACE and TG-43 in order to identify if ACE provided a clinically significant difference in patient dosimetry and could account for inhomogeneity unlike TG-43 if bolus is omitted. Materials and Methods These 27 plans were separated into 2 treatment size areas; small (n=16) and large (n=11). Small areas were classified as treatments that contained fewer than 150 dwell positions such as noses and fingers with an equivalent area of approximately 5cm by 5cm or less. Large areas were classified as treatments that contained 151 or more dwell positions, including limbs and whole scalps. Each plan was recalculated with ACE using the original treatment plan parameters. For the ACE calculations, the patient contour, Perspex immobilisation mask and wax source stand-off were delineated as an external contour and assigned a uniform density of water (ACE assumes the volume outside this external contour to be air). Each plan was recalculated with the original dwell times remaining constant. The average dose to multiple normalisation points situated at the skin surface within the target area for each plan was compared for both TG-43 and ACE calculations. A difference of 5% was determined to be clinically significant. Results The average dose difference (ACE - TG-43) was -3.15% ±0.62% (to one standard deviation) for the small sites. On average, the large treatment sites showed a clinically significant difference in dose of -6.60% ±1.44% when recalculated with ACE. The largest difference between ACE and TG-43 was a -10.57% ±1.44% reduction in dose from TG-43 for a whole scalp treatment. For all plans recalculated ACE predicted a lower superficial dose than TG-43 for the same treatment plan The results suggest ACE was able to account for the lack of backscatter present from the tissue-air interface near to the source dwell positions. The effect was more noticeable in the larger treatment cases, which is likely a result of the larger backscatter reduction and the higher scatter/primary ratio in these larger areas. Further research will involve an experimental verification by end-to-end measurements for clinically relevant scalp and nose treatments. PO-1841 Assessment of three software systems for the independent calculation of Eclipse HyperArc SRS plans J. Calvo-Ortega 1 , M. Hermida-López 2 , C. Laosa-Bello 1 , S. Moragues-Femenia 1 , J. Casals-Farran 1 1 hospital Quirónsalud Barcelona, Radiation Oncology, Barcelona, Spain; 2 hospital Vall D'hebron, Servei De Física I Protecció Radiològica, Barcelona, Spain Purpose or Objective To perform independent dosimetric check of Eclipse HyperArc (HA) SRS plans by using three different software, in the context of patient-specific quality assurance (PSQA). parameters. Conclusion