It is shown that under certain conditions the filtered backprojection algorithm produces a computed tomographic reconstruction for which the statistical accuracy attainable in the amplitude estimation of large-area objects meets the general lower bound derived by Tretiak. In this sense, filtered back-projection is an optimum algorithm.
Keywords: Computed tomography, filtered backprojection, Cramer-Rao lower bound, Fisher information, statistical efficiency, signal-to-noise ratio, image reconstruction, image quality
Get full paper (pdf, 570 KB)
Return to publication list
Send e-mail to author at kmh@hansonhub.com