Abstract
{\it Rational and objectives.} The recent development of image intensifier-based volume tomographic digital angiography has attracted interest in the development of an effective correction protocol for image distortion introduced by an image intensifier (II)-based image chain. This paper presents an effective II distortion-correction method for II-based volume tomographic digital angiography (VTDA).
{\it Methods.} In this calibration protocol, two types of predominant distortion: Global variable magnetic field (GVMF) distortion and pincushion distortion were corrected separately using global mapping functions. GVMF distortion-correction parameters were determined by measuring the translational and rotational parameters from the projection images of a specially designed phantom acquired at different gantry angular positions. A new pincushion distortion-correction algorithm was then developed which remapps projection data acquired with the curved detector onto an imaginary flat detector. The corrected projection data were used for direct 3D reconstruction with a cone-beam reconstruction algorithm for a flat detector. The method was validated through 3D reconstructions of a vascular phantom, a specially designed resolution phantom and animal studies using the newly built II-based volume tomographic digital angiography imaging system.
{\it Results.} Artifacts caused by II distortion in reconstructed images can be completely removed using the II distortion-correction technique. With the distortion-correction, the geometric reconstruction accuracy is significantly improved and the maximum error in vessel diameter is reduced from 2.0 mm without distortion-correction, to 0.3 mm. Moreover, after the distortion-correction, the spatial resolution of the system becomes uniform. The results of 3D reconstruction from phantom and animal studies indicate that the accuracy of the distortion-correction technique is acceptable for volume tomographic digital angiography reconstruction.
{\it Conclusion.} The distortion-correction technique is efficient and provides acceptable accuracy for volume tomographic digital angiography imaging.