ESTRO 2023 - Abstract Book

S24

Saturday 13 May

ESTRO 2023

Alexander’s NTCP. AUCs were 0.61 (CI +-0.04) in both validations; this index cannot be compared with the development cohorts since it was not reported in publications.

Conclusion Two NTCP models for moderate/severe radiation-induced fibrosis were externally validated with very good results on a group of patients with RT regimens analogous to the ones of the development cohorts. BEUD based on the dose distribution in the PTV was a robust predictor of the risk for tissue fibrosis. Data are available to test the transferability of the models outside of the considered RT schedules, including the ultrashort breast irradiation and dose boosts. [Study funded by the EU FP7 programme for research and ERAPerMed]REQUITE was funded from the European Union's 7th FP GA 601826. RADprecise was funded by the ERA PerMed Network, Reference Number ERAPERMED2018-244. MO-0057 Predicting survival outcomes for patients with colorectal cancer by explainable pathology predictors S. Chen 1 , L. Wee 1 , A. Dekker 1 , J. Wang 2 1 Maastricht University Medical Centre+, Department of Radiation Oncology (Maastro), GROW School for Oncology and Reproduction, Maastricht, The Netherlands; 2 Fudan University Shanghai Cancer Centre, Department of Radiotherapy, Shanghai, China Purpose or Objective Although whole slide images (WSI) are the gold standard for tumor grading and postoperative radiotherapy for patients with colorectal cancer (CRC), traditional predictors are usually subjective and machine learning predictors are always lacking interpretability. We proposed an approach to extract explainable and robust predictors from WSIs. Materials and Methods Total 1155 WSIs from 606 patients in The Cancer Genome Atlas (TCGA) – colon adenocarcinoma (COAD) and rectum adenocarcinoma (READ) public dataset and 108 WSIs from 106 patients in local dataset were enrolled as discovery dataset and external validation dataset, respectively. The visual characteristics of tumor gland formation (G), stroma (S), necrosis (N) and tumor-infiltrating lymphocytes (TILs, L) were investigated as prognostic features, which were predicted by deep learning classifiers. The global average or statistic values (minimal, median and maximum) of WSI’s sub-regions were calculated as candidate explainable pathology features (EPFs). The process of EPFs extraction can be seen in Fig.1. Labels of EPFs followed the format of [data source][sub-region level]_[statistic type], for example, the EPF of G4_min referred to the minimal tumor gland formation characteristics at sub-region level of 4. Then the LASSO-embedded features selection was performed. Finally, we developed the Cox regression model on selected EPFs for overall survival evaluation and projected the EPFs on original WSIs.