ESTRO 2022 - Abstract Book

S1592

Abstract book

ESTRO 2022

specific task. Till now, no one has trained a model for RP tasks. In this way, we re-train the pre-train model to extract DL features to make the extracted features suit the RP task. Materials and Methods CT images and dose files were obtained on 349 patients diagnosed with lung cancer who received radical radiotherapy at Tianjin cancer hospital between 2013 to 2019. Compared with the pre-trained model trained on the UCF101 video dataset, we used a linear DL classifier with the cross-entropy loss function to re-train the model. Two experiments were used to prove the features created by re-trained CNN are more suitable for this task than the pre- trained CNN. First, the Principal Component Analysis (PCA) method was used to show the distribution of the re-trained features and pre-trained features. Second, we separate features into training and testing set as 7:3 rate and built a classification model based on logistic regression trained and tested by retrained features and pre-trained features, respectively, then AUC (under the curve) was used to compare the classification performance of the features. Results First, the PCA method shows that the re-trained features have significant differences between positive and negative samples than pre-trained features. Second, compare with the classification model trained and tested by pre-trained features with an AUC score was 0.58, the re-trained features have a higher score of 0.65.

Conclusion Both the pre-trained model and re-trained model have the potential to predict radiation pneumonitis, however, a re-trained model is preferable.

PO-1785 Secondary cancer risk estimates from proton arc plans in pediatric craniopharyngioma.

L. Toussaint 1 , D.J. Indelicato 2 , J.B. Petersen 3 , C.H. Stokkevåg 4 , Y. Lassen-Ramshad 1 , A. Vestergaard 1 , L.P. Muren 1

1 Aarhus University Hospital, Danish Centre for Particle Therapy, Aarhus N, Denmark; 2 University of Florida Health Proton Therapy Institute, Department of Radiation Oncology, Jacksonville, USA; 3 Aarhus University Hospital, Department of Medical Physics, Aarhus N, Denmark; 4 Haukeland University Hospital, Department of Oncology and Medical Physics, Bergen, Norway Purpose or Objective In the past years, interest has been growing around the technological developments and dose tailoring benefits of proton arc therapy (PAT). One of the main concerns for the clinical application of PAT, especially when considering pediatric patients, is the potential enhanced risk of secondary cancer (SC) induction. Indeed, the low-dose bath is increased with PAT compared to state-of-the-art intensity modulated proton therapy (IMPT), and the impact of low doses on SC risks is still unclear. While a few studies investigated SC risk after passive-scattering PAT, reports on active-scanning PAT estimates are absent. The aim of this study was therefore to compare SC risks in pediatric craniopharyngioma patients if treated with either photon-based volumetric modulated arc therapy (VMAT), IMPT or active-scanning PAT. Materials and Methods Treatment plans optimized in Eclipse with a prescribed dose of 54Gy(RBE) were generated on CT-scans from five pediatric craniopharyngioma patients using VMAT, IMPT and PAT-surrogate. The VMAT plans consisted of three arcs, the IMPT plans were calculated with three fields (right/left superior anterior oblique fields, and a superior posterior oblique field), and the PAT plans were arc surrogates plans, i.e. calculated with 18 equiangular beams with a minimal spot weighting of