Key points are not available for this paper at this time.
The advent of the x-ray flat panel imager (FPI) is making the study of cone beam volume CT (CBVCT) more competitive. Motivated by recent encouraging developments in CBVCT, this paper investigates the influence of x-ray scatter on the imaging performance of an x-ray FPI based CBVCT prototype. The prototype employs a circle-plus-two-arc orbit to meet the data sufficiency condition, and can reconstruct a region of interest within a longitudinally unbounded object using a cone beam filtered back-projection algorithm derived for the data acquisition orbit. First, the humanoid phantom is used to investigate the temporal variation of both scatter intensity and scatter to primary ratio (SPR) in the projection images acquired for CB reconstruction. Second, a 160 mm cylindrical water phantom consisting of four 16 mm rods made up of Acrylic, Polyethelene, Polycarborate and Polystrene respectively is utilized to evaluate the variation of interference caused by x-ray scatter (cupping effect) and signal to noise ratio vs. SPR in projection images. Third, a disc phantom consisting of seven acrylic discs stacked at even intervals is employed to evaluate the influence of x-ray scatter on reconstruction accuracy and the improvement of CBVCT image quality with recourse to an anti-scatter grid. Finally, the alleviation of the cupping effect in the presence of a beam-shaping (bow-tie) attenuator is assessed . The quantitative investigation shows that the influence of x-ray scatter on the SNR and CT number accuracy is a crucial problem to be addressed for the application of x-ray CBVCT.
Tang et al. (Thu,) studied this question.