Premium
Filtered‐backprojection reconstruction for a cone‐beam computed tomography scanner with independent source and detector rotations
Author(s) -
Rit Simon,
Clackdoyle Rolf,
Keuschnigg Peter,
Steininger Philipp
Publication year - 2016
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4945418
Subject(s) - imaging phantom , scanner , detector , iterative reconstruction , offset (computer science) , image quality , computer vision , cone beam computed tomography , image guided radiation therapy , field of view , projection (relational algebra) , computer science , artificial intelligence , rotation (mathematics) , optics , algorithm , physics , medical imaging , image (mathematics) , computed tomography , medicine , radiology , programming language
Purpose: A new cone‐beam CT scanner for image‐guided radiotherapy (IGRT) can independently rotate the source and the detector along circular trajectories. Existing reconstruction algorithms are not suitable for this scanning geometry. The authors propose and evaluate a three‐dimensional (3D) filtered‐backprojection reconstruction for this situation. Methods: The source and the detector trajectories are tuned to image a field‐of‐view (FOV) that is offset with respect to the center‐of‐rotation. The new reconstruction formula is derived from the Feldkamp algorithm and results in a similar three‐step algorithm: projection weighting, ramp filtering, and weighted backprojection. Simulations of a Shepp Logan digital phantom were used to evaluate the new algorithm with a 10 cm‐offset FOV. A real cone‐beam CT image with an 8.5 cm‐offset FOV was also obtained from projections of an anthropomorphic head phantom. Results: The quality of the cone‐beam CT images reconstructed using the new algorithm was similar to those using the Feldkamp algorithm which is used in conventional cone‐beam CT. The real image of the head phantom exhibited comparable image quality to that of existing systems. Conclusions: The authors have proposed a 3D filtered‐backprojection reconstruction for scanners with independent source and detector rotations that is practical and effective. This algorithm forms the basis for exploiting the scanner's unique capabilities in IGRT protocols.