303. Radiomic feature changes on cone-beam CT during RT for rectal cancer. (December 2018)

• Record Type: Journal Article
• Authors: Pirrone, G.; Palazzari, E.; De Paoli, A.; Stancanello, J.; Franchin, G.; Farina, E.; Navarria, F.; Cappelletto, C.; Sartor, G.; Avanzo, M.
• Is Part Of: Physica medica. Volume 56(2018)Supplement 2
• Page Start: 247
• Page End: 248
• Publication Date: 2018-12
• Subjects: Medical physics -- Periodicals; Biophysics -- Periodicals; Biophysics -- Periodicals; Imagerie médicale -- Périodiques; Radiothérapie -- Périodiques; Rayons X -- Sécurité -- Mesures -- Périodiques; Physique -- Périodiques; Médecine -- Périodiques; 610.153
• DOI: 10.1016/j.ejmp.2018.04.312
• Languages: English
• ISSNs: 1120-1797

Abstract

Abstract : Purpose: The study of the change of radiomic features extracted from cone-beam CT acquired before each fraction for IGRT opens a new, promising scenario for assessing tumor response to radiotherapy during the treatment[1] . In this preliminary study, we investigated if radiomic features of rectal cancer treated with VMAT extracted from CBCT change significantly in the course of the radiotherapy treatment. Methods and materials: 8 patients underwent VMAT for rectal cancer with 54 Gy in 25 fractions. For each patient, the GTV was manually contoured on three CBCT images acquired for IGRT at the first fraction, middle (12th) and last fraction of RT. All the CBCT images were acquired with 125 kVp and 80 mA with a Varian Clinac Trilogy, and the intensity levels of CBCT images were equalized using the mean intensity level of a ROI in the bladder as reference. Radiomic features were extracted after image denoising and smoothing from 3D tumor regions, and can be classified into: shape-based, histogram-intensity and higher-order texture features. The feature changes were evaluated between last and mid-fraction (Delta1), and between last and first fraction (Delta2). Wilcoxon signed-rank test was used to test for significant changes of features. Results: 18 features had significant Delta1 and 23 had Delta2, while 12 had both changes of statistical relevance (p < 0.05), as shown in Figure-1. The features with lowest p-values were: compactness, surface-to-volume ratio for shape-based features, variance, mean absolute deviation, energy for histogram-intensity features, and coarseness, zone-size variance for texture features. Conclusions: The results exhibit significant changes of radiomic features extracted from daily CBCT images at different treatment times, showing the possibility for studying early response of tumor during therapy and develop a radiomics-adapted radiotherapy approach. The correlation of feature changes with tumor response at the end of therapy will be subject of ongoing investigation.