ESTRO 36 Abstract Book

S866 ESTRO 36 _______________________________________________________________________________________________

radiation track structure. We suggest that this is as a direct consequence of the complexity of the breaks caused, as similar trends are observed for both repair and break induction. This is of relevance for potential application to LET based treatment plans. EP-1605 Deep learning of radiomics features for survival prediction in NSCLC and Head and Neck carcinoma A. Jochems 1 , F. Hoebers 1 , D. De Ruysscher 1 , R. Leijenaar 1 , S. Walsh 1 , B. O'Sullivan 2 , J. Bussink 3 , R. Monshouwer 3 , R. Leemans 4 , P. Lambin 1 1 MAASTRO Clinic, Radiotherapy, Maastricht, The Netherlands 2 Princess Margaret Cancer Centre, Cancer Clinical Research Unit, Toronto, Canada 3 Radboud University Medical Center Nijmegen, Radiation Oncology, Nijmegen, The Netherlands 4 VU University Medical Center, Department of Otolaryngology/Head and Neck Surgery, Amsterdam, The Netherlands Purpose or Objective In order to facilitate personalized medicine in cancer care, predictive models are of vital importance. Radiomics, the high-throughput extraction of large amounts of image features from radiographic images, facilitates predictive model development by providing non-invasive biomarkers. Previous work indicates that radiomics features have high predictive quality 1 . However, these studies used conventional models and the added value of deep learning combined with radiomics features is unexplored. Furthermore, conventional modelling strategies require a selection of features to establish a signature whereas deep learning algorithms do not. In this work we learn a deep learning model on radiomics features and compare it to a previously published cox regression model 1 . Material and Methods 4 independent Lung and Head & Neck (H&N) cancer cohorts (1418 total patients) were used in this study. Radiomic features were extracted from the pre-treatment computed tomography images. The model was learned on the Institute 1 lung cohort (N=422) and validated on the other datasets. The outcome is two-year survival following treatment. A 3 layer deep learning network was used to make predictions. Results Validation on Institute 2 dataset (N=154) yields an AUC of 0.71 (95% CI: 0.63-0.8) for the deep learning network and 0.66 on the conventional model (95% CI: 0.56-0.75). The difference is not significant (P=0.11). Validation on Institute 3 dataset (N=95) yields an AUC of 0.64 (95% CI: 0.53-0.79) for the deep learning network and 0.75 on the conventional model (95% CI: 0.64-0.86). The difference is not significant (P=0.19). Validation on Institute 4 dataset (N=136) yields an AUC of 0.71 (95% CI: 0.59-0.8) for the deep learning network and 0.74 on the conventional model (95% CI: 0.64-0.83). The difference is not significant (P = 0.24). Validation on Institute 5 dataset (N=540) yields an AUC of 0.58 (95% CI: 0.52-0.63) for the deep learning network and 0.65 on the conventional model (95% CI: 0.59- 0.70). The difference is not significant (P = 0.10).

Figure 1: ROC curves of the model validation. Conclusion

The combination of deep learning and radiomics features has similar performance to conventional radiomics modelling strategies. However, feature selection is no longer a required component as all features can be included in the network. This is a major advantage as feature selection is a computationally intractable task for which only heuristic solutions exist.

References 1 Aerts, H. et al, Nat. Commun. 2014 , 5 , 4006.

EP-1606 Calculating ion-induced cell death and chromosome damage by the BIANCA biophysical model M.P. Carante 1,2 , F. Ballarini 1,2 1 Istituto Nazionale di Fisica Nucleare INFN, Section of Pavia, Pavia, Italy 2 University of Pavia, Physics Department, Pavia, Italy Purpose or Objective To calculate probabilities of cell death and chromosome aberrations following cell irradiation with ion beams of different energy. Material and Methods A biophysical model called BIANCA (BIophysical ANalysis of Cell death and chromosome Aberrations) [Carante M.P. and Ballarini F. Front. Oncol. 6:76 2016] was refined and applied to simulate cell death and chromosome aberrations by therapeutic protons and heavier ions. The model, which assumes a pivotal role for DNA cluster damage, is based on the following assumptions: i) a DNA “Cluster Lesion” (CL) produces two independent chromosome fragments; ii) chromosome fragment un- rejoining, or distance-dependent mis-rejoining , gives rise to chromosome aberrations; iii) certain aberrations (dicentrics, rings and large deletions) lead to cell death. The CL yield is an adjustable parameter, as well as the probability that a chromosome fragment remains un- rejoined even if possible partners for rejoining are present. The model, implemented as a MC code providing simulated dose-response curves comparable with experimental data, was applied to different beams, including beams available at the CNAO hadrontherapy centre in Pavia, Italy, and at the CATANA facility in Catania, Italy. Results The model allowed reproduction of experimental survival curves for cell lines characterized by different radiosensitivity, supporting the model assumptions. Furthermore, cell death and chromosome aberrations