Brachytherapy seed reconstruction with joint-encoded C-arm single-axis rotation and motion compensation

C-arm fluoroscopy images are frequently used for qualitative assessment of prostate brachytherapy. Three-dimensional seed reconstruction from C-arm images is necessary for intraoperative dosimetry and quantitative assessment. Seed reconstruction requires accurately known C-arm poses. We propose to measure the C-arm rotation angles and computationally compensate for inevitable C-arm motion to compute the pose. We compensate the translational motions of a C-arm, such as oscillation, sagging and wheel motion using a three-level optimization algorithm and obviate the need for full pose tracking using external trackers or fiducials. We validated our approach on simulated and 100 clinical data sets from 10 patients and gained on average, a seed matching rate of 98.5%, projection error of 0.33 mm (STD = 0.21 mm) and computation time of 19.8 s per patient, which must be considered as clinically excellent results. We also show that without motion compensation the reconstruction is likely to fail.

Motion compensated seed reconstruction.Figure optionsDownload high-quality image (111 K)Download as PowerPoint slideHighlights
► During brachytherapy, we acquire C-arm images by planar rotation of the C-arm. C-arm rotation angles are measured using protractors or joint encoders.
► We computationally compensate for C-arm motion in 2D space, using the seeds.
► We obviate the need for full pose tracking using fiducials or external trackers.
► We achieved on average 98% matching rate and 0.3 mm projection error per patient.