The SV registration had a precision of 2.7° (1 standard deviation) and was successful for 96% of the scans. Mean (M), systematic (Σ), and random (σ) differences between the orientation of the prostate and SV were M = -0.4°, Σ = 7.2°, and σ = 6.4°. Daily marker-based corrections required an SV margin of 11.4 mm (translations only) and 11.6 mm (translations + rotations). Rotation corrections of the SVs reduced the required margin to 8.2 mm.
Twenty prostate patients implanted with 2 or 3 elongated fiducial markers had cone beam computed tomography (CBCT) scans acquired at every fraction. The first step of the hybrid registration was to localize the prostate by CBCT-to-planning-CT alignment of the fiducial markers, allowing both translations and rotations. Using this marker registration as a starting point, the SVs were registered based on gray values, allowing only rotations around the lateral axis. We analyzed the differential rotation between the prostate and SVs and compared the required SV margins for 3 correction strategies.
We found substantial differences between the orientation of the prostate and SVs. The hybrid registration technique can accurately detect these rotations during treatment. Rotation correction of the SVs allows for margin reduction for the SVs.
Fiducial markers are a good surrogate for the prostate but provide little information on the position and orientation of the seminal vesicles (SVs). Therefore, a more advanced localization method is warranted if the SVs are part of the target volume. The purpose of this study was to develop a hybrid registration technique for the localization of the prostate and SVs.
This website uses cookies to ensure you get the best experience on our website.