Room: ePoster Forums
Purpose: To compare the reproducibility of deep-inspiration breath-hold (DIBH) monitored using 3D surface imaging (AlignRT, VisionRT, UK) and infrared marker tracking (RPM, Varian, CA) systems.
Methods: Two groups of 10 patients were considered. DIBH was monitored using AlignRT for the first group, and using RPM for the second group. All patients were positioned by aligning on skin tattoos at treatment. DIBH was monitored by matching to the patient surface from CT scans for the AlignRT group, and by tracking the infrared markers for the RPM group. To align the bony landmarks for position verification, daily orthogonal kV images were acquired while patients were performing DIBH. Couch shifts in 6 degrees-of-freedom (DOF) were applied to correct setup errors from breath-hold. Potential setup errors from DIBH variability, without daily kV images, were assessed by evaluating 6DOF couch shifts for the patients. The image-guided workflow efficiency was assessed by counting the number of repeated orthogonal kV images for each fraction. Population-based PTV margins considering DIBH variability were estimated for both systems.
Results: A total of 262 and 258 fractions for the AlignRT group and RPM group were evaluated, respectively. The mean and standard deviation of the number of kV image pairs for each fraction was 1.12Â±0.34 and 1.30Â±0.53 for AlignRT and RPM respectively (p<0.05 for t-test). The PTV margins for superior-inferior (S/I) and left-right (L/R) directions were 0.8cm and 0.5cm for AlignRT, and 1.3cm and 1.2cm for RPM. However, the PTV margins for the anterior-posterior (A/P) direction were 0.8cm for AlignRT and 0.9cm for RPM. All the rotations were comparable, the largest being 2.7â?° in the pitch direction.
Conclusion: Breath-hold was more reproducible in S/I and L/R when monitored with AlignRT than RPM. However, DIBH variability requires daily orthogonal kV images to correct setup errors, or adequately expanded PTV margins in treatment planning.