Poster ‐ 57: Statistical analysis of setup correction for radical IMRT lung patients on a treatment couchtop with six degrees of freedom

Abstract

Purpose: To characterize the distributions of setup corrections for radical IMRT lung patients treated on a couchtop with six degrees of freedom (6DOF). Methods: Translational and rotational setup corrections were retrospectively analyzed for patients treated on the HexaPODTM 6DOF couchtop. Conventional and hypo‐fractioned radical IMRT lung treatments were included, for a total of 152 patients and 2,042 fractions. The distribution of setup corrections was analyzed for each DOF. Potential correlations between pairs of corrections were examined and possible causes identified. Intra‐patient variability of setup corrections was characterized. Results: For lateral, longitudinal and vertical setup corrections, the mean ± 1 S.D. (in mm) were: −0.6 ± 3.8, 2.8 ± 5.1, and 1.1 ± 3.9. Each of the three means is statistically significant from the theoretical population mean of zero (p<0.01). For rotations, the means were within 0.1° from zero, with a standard deviation of 1.3°. Moderate correlation (r = −0.45) was observed between longitudinal and pitch corrections, which can be caused by changes in the spinal column position, to which registration is mostly done. The standard deviation for intra‐patient corrections varies among patients: 2 – 10 mm for translations and 0.5 – 3° for rotations. Intra‐patient variability in translations has a positive correlation with that in rotations. Conclusions: In 6DOF setup corrections for lung patients, systematic translational corrections were observed, with moderate correlation between longitudinal and pitch corrections. Careful setups and custom immobilization may help reduce such offsets, correlations, and intra‐patient setup variability.