Measuring interfraction and intrafraction lung function changes during radiation therapy using four‐dimensional cone beam CT ventilation imaging

• DOI: 10.1118/1.4907991
• Publication Date: 2015-02-20T18:00:30Z
• Authors: John Kipritidis; Geoffrey Hugo; Elisabeth Weiss; Jeffrey Williamson; Paul J. Keall
• Keywords: Cone-Beam CT; Lung ventilation; Image-guided radiation therapy

Adaptive ventilation guided radiation therapy could minimize the irradiation of healthy lung based on repeat imaging (VI) during treatment. However efficacy adaptive guidance requires that interfraction (e.g., week-to-week), changes are not washed out by intrafraction pre- and postfraction) changes, for example, due to patient breathing variability. The authors hypothesize patients undergoing cancer exhibit larger compared function changes. To test this, perform first comparison VI pairs using four-dimensional cone beam CT (4D-CBCT VI), a novel technique functional imaging.The analyzed total 215 4D-CBCT scans acquired 19 locally advanced non-small cell (LA-NSCLC) over 4-6 weeks therapy. This set was sorted into 56 (including day each treatment 2, 4, 6) 78 pre/postfraction same-day), with some appearing in both sets. VIs were obtained from Jacobian determinant transform between end-exhale end-inhale images after deformable image registration. All deformably registered their corresponding planning normalized account differences effort, thus facilitating terms (i) voxelwise Spearman correlations, (ii) mean differences, (iii) gamma pass rates all pairs. For side ipsilateral tumor, we applied two-sided t-tests determine whether more different than pairs.The (mean ± standard deviation) correlation r̄(Inter)=0.52±0.25, which significantly lower (r̄(Intra)=0.67±0.20, p = 0.0002). Conversely, absolute pairs, |ΔV̄(Inter)|=0.42±0.65 |ΔV̄(Intra)|=0.32±0.53, respectively (p < 10(-15)). Applying analysis ventilation/distance tolerance 25%/10 mm, observed rate (69% 20%) VIs, (80% 15%, ∼ 0.0003). Compared scans, experienced at least one subsequent change median ≥10%. Patients positive negative throughout treatment, maximum occurring patients.The authors' data support hypothesis LA-NSCLC course conventional Longitudinal be highly patient-dependent, supporting possible role VIs. We anticipate future improvement reconstruction algorithms will improve capability resolve