Abstract Book

ESTRO 37

S551

using IBEX. Spearman correlation was used to reduce the 145 features to 7 (cutoff: 0.7). Features were: ‘5-7 Correlation’, ‘11-7 Correlation’, ‘9-7 Correlation’ from GLCM3 category,’90-7 Correlation’ from GLCM2.5 category, ‘Skewness’ from ‘GradientOrientHistogram’ Category, ‘Busyness’ from NID2.5, and ‘Orientation’ from ‘Shape’ category. These were integrated to build 3 types of models using: 1) only the baseline (BL) value of the radiomics features, 2) the ratio between the mid- treatment value of the feature to its value at BL and 3)a functional principal component analysis model leveraging the structure of the temporal trajectory in the evolution of the feature from BL to mid-treatment. Afterwards, logistic regression was run to investigate the predictive capacity of clinical attributes as well as the synergistic effect of adding clinical attributes & either one of the three radiomic-derived models. Results 39 predominately locally advanced OPC patients who received concurrent chemoradiotherapy were included. 24 patients (61.5%) achieved complete response (CR) on the primary tumor, as compared to 15 (38.5%) partial responses (PR). The corresponding areas under the curve (AUCs) and confidence intervals (C.I.) for the prediction of tumor response to (chemo)radiation ‘CR vs PR’, according to the 3 models, augmented by clinical prognosticators, were: 0.44 [95%C.I.:0.25-0.64], 0.64 [95%C.I.:0.42-0.8] and 0.71 [95%C.I.:0.52-0.87], respectively. Whereas, the clinical attributes only model yielded a lower AUC of 0.49 [95%C.I.:0.31-0.7]. This suggests the functional approach yields a superior predictive power in the evaluation of tumor response compared to other radiomic models or even strictly clinical models. ( Figure 1 )

Figure 1: The color of the mesh indicates the distance error between the appearance model and the ground truth. Left: Iteration 5, differences 491; Middle: Iteration 10, differences 311; Right: Iteration 21, differences 197. 1 2 3 4 5 6

Clinical Vol (cm3)

45

42

28

87

65

58

Model Vol (cm3)

44

45

32

77

65

62

Dice

Similarity

0.88 0.77 0.84 0.79 0.93 0.89

Coefficient

Table 1 : Dice coefficient and volume obtained by the appearance model. Conclusion The proposed model has the potential to be used for automatically contouring the GTV after brachytherapy treatment. However, the 10 pixel margin used to extract the appearance patch needs to be further investigated to fit smaller GTV volumes. In addition with further work the approach has the potential to be used for tracking disease progression. PO-0991 Serial tumor radiomic features predict response of head and neck cancer treated with Radiotherapy H.E. Elhalawani 1 , A.S.R. Mohamed 1 , S. Volpe 1 , P. Yang 1 , S. Campbell 1 , R. Granberry 1 , R. Ger 1 , X. Fave 1 , L. Zhang 1 , G.E. Marai 2 , D. Vock 3 , G.M. Canahuate 4 , D. Mackin 1 , L. Court 1 , G.B. Gunn 1 , A. Rao 1 , C.D. Fuller 1 1 The University of Texas- MD Anderson Cancer Center, Radiation Oncology, Houston, USA 2 University of Illinois at Chicago- Chicago- Illinois- USA., Computer Science, Chicage, USA 3 University of Minnesota of Public Health- Minneapolis- Minnesota- USA., Biostatistics, Minneapolis, USA 4 University of Iowa- Iowa City- IA- USA, Electrical & Computer Engineering, Iowa City, USA Purpose or Objective Predicting ultimate tumor response before/during radiotherapy (RT) is key for risk stratification and subsequent treatment individualization. We analyzed radiomic features longitudinally for quantifying changes in tumoral structure in a cohort of head and neck cancer (HNC) patients. We studied how clinical and temporally- derived imaging features can be integrated into a multifactorial predictive tool of treatment outcome. Material and Methods HNC patients undergoing image-guided RT were included. Primary tumor response at the end of RT course per RECIST v1.1 was retrieved. Baseline patient and disease characteristics were recorded. A total of 155 in- treatment CT scans at days 1, 5, 10 and 15 were reclaimed. Primary gross tumor volumes were contoured. A total of 145 radiomics features were selected from the categories: intensity direct, neighborhood intensity difference (NID), grey-level co-occurrence matrix (GLCM), grey-level run length and shape, and analyzed

Conclusion Temporal radiomic trajectories from sequential intra- treatment CT scans can be integrated with traditional clinical attributes into a multi-faceted decision-making tool. Radiomics may convey additional clinical information from routine imaging studies in radiation oncology towards adaptive RT.

Poster: Physics track: Implementation of new technology, techniques, clinical protocols or trials (including QA and audit)

PO-0992 Electron IORT in vivo film dosimetry in breast cancer for alignment and dose verification A. Petoukhova 1 , J. Nijst-Brouwers 1 , K. Van Wingerden 1 , J. Van Egmond 1 , T. Stam 2 , A. Marinelli 3 , J. Van der Sijp 3 , M. Straver 3 , O. Guicherit 3 , P. Koper 2 , H. Struikmans 4