RESUMO
PURPOSE: To assess the interobserver delineation variability of radiomic features of the parotid gland from computed tomography (CT) images and evaluate the correlation of these features for head and neck cancer (HNC) radiotherapy patients. MATERIALS AND METHODS: Contrast-enhanced CT images of 20 HNC patients were utilized. The parotid glands were delineated by treating radiation oncologists (ROs), a selected RO and AccuContour auto-segmentation software. Dice similarity coefficients (DSCs) between each pair of observers were calculated. A total of 107 radiomic features were extracted, whose robustness to interobserver delineation was assessed using the intraclass correlation coefficient (ICC). Pearson correlation coefficients (r) were calculated to determine the relationship between the features. The influence of excluding unrobust features from normal tissue complication probability (NTCP) modeling was investigated for severe oral mucositis (grade ≥3). RESULTS: The average DSC was 0.84 (95% confidence interval, 0.83-0.86). Most of the shape features demonstrated robustness (ICC ≥0.75), while the first-order and texture features were influenced by delineation variability. Among the three observers investigated, 42 features were sufficiently robust, out of which 36 features exhibited weak correlation (|r|<0.8). No significant difference in the robustness level was found when comparing manual segmentation by a single RO or automated segmentation with the actual clinical contour data made by treating ROs. Excluding unrobust features from the NTCP model for severe oral mucositis did not deteriorate the model performance. CONCLUSION: Interobserver delineation variability had substantial impact on radiomic features of the parotid gland. Both manual and automated segmentation methods contributed similarly to this variation.
RESUMO
Modern radiotherapy (RT) uses altered fractionation, long beam-on time and image-guided procedure. This study aimed to compare secondary cancer risk (SCR) associated with primary field, scatter/leakage radiations and image-guided procedure in prostate treatment using intensity-modulated RT (IMRT), CyberKnife stereotactic body RT (CK-SBRT) in relative to 3-dimensional conformal RT (3D-CRT). Prostate plans were generated for 3D-CRT, IMRT (39 fractions of 2 Gy), and CK-SBRT (five fractions of 7.25 Gy). Excess absolute risk (EAR) was calculated for organs in the primary field using Schneider's mechanistic model and concept of organ equivalent dose (OED) to account for dose inhomogeneity. Doses from image-guided procedure and scatter/leakage radiations were determined by phantom measurements. The results showed that hypofractionation relative to conventional fractionation yielded lower SCR for organs in primary field (p ≤ 0.0001). SCR was further modulated by dose-volume distribution. For organs near the field edge, like the rectum and pelvic bone, CK-SBRT plan rendered better risk profiles than IMRT and 3D-CRT because of the absence of volume peak in high dose region (relative risk [RR]: 0.65, 0.22, respectively, p ≤ 0.0004). CK-SBRT and IMRT generated more scatter/leakage and imaging doses than 3D-CRT (p ≤ 0.0002). But primary field was the major contributor to SCR. EAR estimates (risk contributions, primary field: scatter/leakage radiations: imaging procedure) were 7.1 excess cases per 104 person-year (PY; 3.64:2.25:1) for CK-SBRT, 9.93 (7.32:2.33:1) for IMRT and 8.24 (15.99:2.35:1) for 3D-CRT (p ≤ 0.0002). We conclude that modern RT added more but small SCR from scatter/leakage and imaging doses. The primary field is a major contributor of risk which can be mitigated by the use of hypofractionation.
