ESTRO 2021 Abstract Book

S427

ESTRO 2021

CNNs generally showed good performance for all classification levels (Table 1), with overfitting becoming slightly more apparent for the more detailed classification levels. For cohort 1, (3%, 1mm) gamma maps provided consistently better results than commonly used (3%, 3mm) gamma maps. As the CNNs used for cohort 3 were optimized for TI data, but not yet for TR data, their performance is lower.

Conclusion Deep learning is a promising powerful tool for identifying types and magnitudes of treatment errors with EPID dosimetry. It can advance the field of EPID dosimetry substantially, by providing rapid, automated models for error identification, and with that, additional detailed information not currently available from EPID dosimetry. Currently, the applicability of these CNNs to clinically measured EPID data, as well as optimization of CNNs for TR data and the use of different gamma criteria and dose comparison methods as model input are under further investigation for these lung cancer cohorts. PH-0544 Clinical impacts of managing motion in kidney stereotactic ablative body radiotherapy M. Gaudreault 1 , S. Siva 2 , T. Kron 1 , N. Hardcastle 1 1 Peter MacCallum Cancer Centre, Physical Sciences, Melbourne, Australia; 2 Peter MacCallum Cancer Centre, Division of Radiation Oncology, Melbourne, Australia Purpose or Objective Kidney stereotactic ablative body radiotherapy (SABR) is a novel technique to treat renal cancer cell (RCC) for medically inoperable patients. The adjacent non-tumour ipsilateral kidney receives substantial dose during SABR treatment, reducing post-treatment renal function. This study aims to quantify dose reduction to the healthy ipsilateral kidney from elimination of respiratory motion and resultant estimated gain in renal function in kidney SABR. Materials and Methods In 62 patients who received kidney SABR, the gross tumour volume (GTV) was segmented on each phase of a respiratory phase-binned 4-dimensional CT (4DCT). Tumour motion was determined from the GTV centroid position on each phase. Motion managed (MM) plans were defined as those where respiratory motion is effectively removed such as that achieved with gating, breath hold or tracking approaches. MM plans were optimized and calculated to MM GTV on the exhale phase image. Internal target volume (ITV) was created based on the full respiratory excursion, and ITV-based plans were optimized and calculated on the average intensity projection of a phase-binned 4DCT. A 5 mm isotropic PTV margin was used in both plans. Low modulation volumetric modulated arc plans were created. To simulate respiratory-induced motion, the ITV- based plan was copied ten times to the exhale image with the isocentre shifted according to the respiratory motion of the tumour relative to the beam. Dose was calculated on each phase and averaged to result in non-motion managed (NMM) plans. Relative change to the ipsilateral healthy kidney volume receiving 50% of the prescription dose (V50%) was assessed as a function of tumour motion amplitude (TMA). Furthermore, gain to renal function resulting from dose reduction was estimated from the change in the glomerular filtration rate (ΔGFR) between MM plan and NMM plan by using a previously published dose-response model derived from kidney SABR [1]. Results The mean ± standard deviation TMA was 0.7 ± 0.5 cm, 0.3 ± 0.2 cm, and 0.2 ± 0.1 cm in the sup-inf, ant- post and left-right directions respectively. A linear relationship between GTV to ITV volume difference and TMA was observed (slope = -26.5 %/cm, Pearson-r = -0.83, p-value < 10 -16 ). A larger volume of healthy ipsilateral kidney was spared in MM plan compared with NMM plan, with difference in absolute volume receiving 50% of prescription isodose increasing linearly with increasing TMA (slope = 12 cc/cm, r = 0.69, p- value < 10 -9 ) as shown in Fig. 1. Consequently, the estimated mean ΔGFR was improved with increasing TMA (slope = 4.4%/cm, r = 0.85, p-value < 10 -10 ). Conclusion Eliminating respiratory motion through MM in kidney SABR reduces the dose received to the non-tumour ipsilateral kidney and consequently reduces chance of renal function loss post SABR treatment. [1] S. Siva, et al., Radiother Oncol, 118, 540–546, 2016.