Abstract Book

S1085

ESTRO 37

planar imaging sequence with b -values of 0- 1000 s/mm 2 , and TR/TE = 3000/120 ms provided data for Apparent Diffusion Coefficient (ADC) maps. 41 lesions were contoured and labeled according to respective patient gleason score. Radiomics features extracted from 2 manual contours and 3 thresholds contouring algorithm (range 120-140% of the minimum value) were correlated with gleason score. A predictive model was developed by selecting features able to discriminate gleason 6 vs higer grade lesions among all the 5 contouring styles(fig1).

Conclusion Majority of radiomics shape features have linear direct or inverse correlations with spiculatedness. We have identified five features that are significantly correlated with spiculatedness and not significantly correlated with volume. However, prudence is still necessary when comparing tumors with differing volumes as features do not appear entirely constant with volume variations. EP-1994 Radiobiological modeling analysis of different fraction schemes for nasopharyngeal cancer J.Y. Lu 1 , B.T. Huang 1 1 Cancer Hospital of Shantou University Medical College, Radiation Oncology, Shantou, China Purpose or Objective The aim of this study was to predict preferable radiotherapy fraction scheme (FS) for early- and advanced-stage nasopharyngeal cancer (NPC), respectively, using radiobiological modeling analysis. Material and Methods 20 cases of early- and advanced-stage NPC were included, respectively. For each case, volumetric modulated arc therapy (VMAT) plans were designed with the same constraints based on four different FSs (70Gy/35 fractions, 70Gy/33 fractions, 70Gy/32 fractions, 70Gy/30 fractions), respectively. Tumor control probability (TCP) was calculated by using the Webb-Nahum model considering tumor repopulation and hypoxia. Normal tissue complication probabilities (NTCPs) were calculated utilizing Lyman-Kutcher-Burman (LKB) model. The four kinds of FSs were compared for early- and advanced-stage NPC, respectively, based on uncomplicated tumor control probability (UTCP), TCP and NTCP analysis. Results For the early stage NPC, the median UTCPs for the FSs of 70Gy/35 fractions, 70Gy/33 fractions, 70Gy/32 fractions, 70Gy/30 fractions were 77.6%, 79.6%, 80.9%, 82.4% ( p < 0.05), respectively, while the median TCPs were 85.9%, 88.5%, 89.8%, 92.8%. The NTCPs of 70Gy/30 fractions FS only increased by 0% -1.6% compared to 70Gy/35 fractions while the NTCPs of most organs at risk were ≤ 5%. The UTCPs of advanced NPC were not significantly different (33.2%, 31.9%, 31.2% and 30.4%, p > 0.05). Conclusion Based on radiobiological modeling analysis, it is recommended to use 70Gy/30 fractions FS to improve the local control rate of tumor while not significantly increase NTCPs for early NPC cases. For advanced NPC, the predicted benefits are similar for different FSs. EP-1995 Robustness of textural features in ADC maps with different contouring for prostate cancer patients G. Feliciani 1 , S. Francesca 2 , S. Barbieri 3 , M. Del Duca 4 , G. Mazzotti 1 , V. Ravaglia 1 , A. Sarnelli 1 , G. Parenti 5 , F. Ansaloni 5 , F. Albarello 5 , E. Menghi 1 1 IRST, Medical Physics, Meldola, Italy 2 University of Ferrara, School of Medicine, Ferrara, Italy 3 University of Ferrara, School of specialization in Radiology, Ferrara, Italy 4 IRST, Radiotherapy, Meldola, Italy 5 AUSL Ravenna, Radiology, Ravenna, Italy Purpose or Objective Among Radiomics studies MR ones are the most challenging. This is due to variability in acquisition and reconstructions methods. In this study we proof robustness of features in prostate cancer patients which predict gleason score from ADC maps employing 5 different contouring methods Material and Methods In our study 25 patients diagnosed with prostate cancer from biopsy and staged by gleason score were enrolled for MR scan and subsequent radiotherapy. A DW echo

Fig1 Different contouring methods, 2 manual contouring, 120%, 130%, 140% threshold contouring of lesion and healthy tissue Results Correlation with gleason 6 versus higher grade was found for one feature for the manual contouring(fig2). Four more feature agreed with the main endpoint employing the automatic contouring in the threshold range. A model based on the feature selected by manual contouring scored an AUC of 0.79 whereas the model based on threshold algorithm an AUC of 0.85.

Fig2 Grey level run length matrix - Short Run Emphasis feature boxplot for 2 manual contours in blue and green. Dashed red line shows the threshold of the model. Conclusion Our study underlined the critical importance in the choice of contouring style. Agreement of 4 feature correlation with gleason score was confirmed for semi automatic algorithm in a wide threshold range scoring a high predictive power. Validation with 8 external patient is ongoing. EP-1996 A patient-specific tumor control probability model based on total lesion glycolysis of anal cancer V. Skingen 1 , E. Rusten 1 , B. Rekstad 1 , C. Undseth 2 , M. Guren 2 , E. Malinen 3 1 Oslo University Hospital, Department of Medical Physics, Oslo, Norway 2 Oslo University Hospital, Department of Oncology, Oslo,